Skip to main content
Look up an error or warning code that the ABL compiler, the parser, or Studio reports. Each entry gives the meaning and the fix. Codes that start with E or W and a number come from the parser. Named codes come from the compiler. An error blocks compile or publish. A warning does not. This page lists the codes that authors meet most often. It is not the full set.

E394

Severity: error Meaning. SET_INTERACTION_LANGUAGE switches the conversation language when the action fires, and its value must be a valid BCP-47 tag. The authored value did not canonicalize to one, so the language switch is dropped while the rest of the action still runs. Fix. Use a valid BCP-47 tag, a language code, optionally with a region suffix, such as en, en-US or pt-BR.

E395

Severity: error Meaning. CLEAR_INTERACTION_LANGUAGE resets the conversation back to the default language and is a bare directive: naming it is the whole instruction. A value after the colon has no meaning, so the parser rejects the line instead of silently discarding what was authored there. Fix. Write the directive as ”- CLEAR_INTERACTION_LANGUAGE:” with nothing after the colon.

E396

Severity: error Meaning. set_interaction_language switches the conversation language when the action fires, and its value must be a string carrying a valid BCP-47 tag. A non-string value or a string that does not canonicalize to a tag drops the language switch while the rest of the action still applies. Fix. Set the key to a valid BCP-47 tag string, a language code, optionally with a region suffix, such as “en”, “en-US” or “pt-BR”.

E492

Severity: error Meaning. FORMAT declared the kind of value a session field holds, and drove kind inference for MATCH_SCORE, MATCHES and MATCH_RESULT. Nothing applied it once DERIVED was removed, so it changed no behaviour. The key is rejected rather than ignored so a file cannot look as if it configures matching when it does not. Fix. Delete the FORMAT line and pass the kind as the third argument of MATCH_SCORE, MATCHES or MATCH_RESULT where the comparison is written.

E493

Severity: error Meaning. SOURCE: tool made a session field writable only by tool bindings and refused a model write. It was removed. The key is rejected rather than ignored so a field that looks write-protected is not silently left writable by the model. Fix. Replace SOURCE: tool with READONLY: true on the field. Authored writes (SET, on_result, hooks) still write a READONLY field.

E498

Severity: error Meaning. DERIVED declared a session field the runtime recomputed after every session write. It was removed because the recompute cost grew with every write. The protection it gave, that the model can never set the field, is what READONLY: true provides, and the value itself is computed by an authored hook SET. The key is rejected rather than ignored so a field that nothing computes cannot compile. Fix. Replace DERIVED: <expression> with READONLY: true on the field, and compute the value with a SET (SET: field = <expression>) in a hook that runs when its inputs change.

E580

Severity: error Meaning. PARALLEL is a block. Write its calls, and in the full form its options, on indented lines below it. A value on the same line is rejected. Fix. Move the value to indented lines, for example PARALLEL: then - CALL: get_balance.

E581

Severity: error Meaning. A PARALLEL block accepts only its own options (WAIT, FAIL_FAST, MAX_CONCURRENCY, TIMEOUT, STATUS_AS, CALLS). Each call accepts only CALL, WITH, AS, WHEN, REQUIRED, TIMEOUT, RETRY, RETRY_DELAY, RETRY_BACKOFF, SET, DEFAULT, ERROR_AS, FALLBACK, AFTER, FOR_EACH, ITEM_AS, INDEX_AS, FILTER, MAX_ITEMS, and KEY_BY. Keys are upper case. An unknown key is rejected so a misspelled control is never ignored. Fix. Check the spelling and case of the key, or remove it.

E582

Severity: error Meaning. Each key can appear once in the PARALLEL options and once per call. A repeat would let one value silently win over the other. Fix. Keep one occurrence of the key. Use the block form of SET for several assignments.

E583

Severity: error Meaning. In the list form, every line under PARALLEL starts a call with - CALL: or continues one. In the full form, options and CALLS sit at one indent, and calls are list items under CALLS. Do not mix list items and options at the same level. Each call key must line up with the CALL key above it. Fix. List the calls directly under PARALLEL, or put the options and a CALLS: list under it. Align each call key under CALL.

E584

Severity: error Meaning. The value does not have the form this PARALLEL key takes. Durations are 1500ms, 2s, or a whole number of milliseconds. Counts are whole numbers. Flags are true or false. DEFAULT is single-line JSON. Names are identifiers. Fix. Write the value in the form quoted in the message.

E585

Severity: error Meaning. Each call in a PARALLEL block is a list item whose first key is CALL with a bare tool name. Inline arguments such as tool(a: b) and tool AS x are not accepted. Fix. Write - CALL: tool_name, then WITH, AS, and the other keys on the lines below it.

E586

Severity: error Meaning. A PARALLEL block must hold at least one call, either as list items or under CALLS. Fix. Add at least one - CALL: tool_name item.

E587

Severity: error Meaning. PARALLEL blocks work only in ABL text. YAML would ignore an unknown step key and run the step without its calls, so the compiler rejects the key. Fix. Convert the agent to ABL text, or move the calls into separate CALL steps.

W170

Severity: warning Meaning. The key is indented under a GATHER field but is not one the field parser reads, so its value is discarded. This used to happen silently, it is how authored EXTRACTION_HINTS vanished, and the warning exists so a misspelt property surfaces at compile time instead of as a field that never extracts. A near-match suggestion is appended when the key looks like a known one. Fix. Correct the key to the property you meant, the suggestion in the message when there is one, or delete the line if it was left over from an earlier shape of the field.

W174

Severity: warning Meaning. A FLOW GATHER field reads VALIDATE as a single regular expression on the same line. A VALIDATE block (type, rule and error lines, or a ”|” / ”>” block scalar) is not read: it is skipped whole, so its lines never become field properties, and it adds no validation. VALIDATION_PROCESS, RETRY_PROMPT and MAX_RETRIES only mean something next to a rule, so the compiler ignores them for a field that, merged with the agent-level GATHER field of the same name, has no rule (no non-blank VALIDATE, and no enum values on a type: enum field). That includes retry settings inherited from the agent-level field. Fix. Put the regular expression on the VALIDATE line itself, and keep RETRY_PROMPT and MAX_RETRIES as field properties: VALIDATE: ”^[0-9]{10}$”, RETRY_PROMPT: ”…”, MAX_RETRIES: 3.

W175

Severity: warning Meaning. A FLOW GATHER field reads VALIDATE as a single regular expression. An inline object such as { pattern: ”…” } or a list is not parsed: the whole text, braces included, becomes the regex, so valid input is rejected or invalid input passes. Fix. Replace the object or list with the regular expression itself: VALIDATE: ”^[0-9]{10}$”.

W208

Severity: warning Meaning. A SECURE_INPUT block accepts MODE, FIELDS, MAX_ATTEMPTS, CANCEL, SAFE_DETAILS_AS, PROGRESS_PROMPT, the delivery CALL with its AS, RESULT_FIELDS and SUCCESS_WHEN, and the four outcome routes. Any other line had no field to land in and was silently discarded. The common case is a RESPOND written under ON_SUCCESS: the route was left empty, so after the payment went through the flow had nowhere to go and the caller heard the empty-response fallback. Fix. Name a FLOW step in the outcome route (ON_SUCCESS: payment_done) and write the RESPOND in that step.

W210

Severity: warning Meaning. A SECURE_INPUT field is read as PROMPT, RETRY_PROMPT, NO_INPUT_PROMPT, COLLECT, VALIDATE, READ_ONCE, PROGRESS_PROMPT, SAVE_AS, the readable-value marker, and the readable-field settings MODE, RECORD and SPEECH_AS (see the SECURE_INPUT reference) only. Any other line had nowhere to land and was silently discarded. The common case is a MESSAGE_KEY written under a field: it was dropped, so the caller heard the PROMPT text (often a placeholder) instead of the localized message the author pointed at. Fix. Reference the locale entry from the prompt itself: PROMPT: “{{locale.PAYMENT_ENTER_CARD_NUMBER}}”, or a language map such as PROMPT: { en: ”…”, es: ”…” }.

W211

Severity: warning Meaning. RESULT_FIELDS and SUCCESS_WHEN are read from the same line only. A list or map written on the following lines left the value empty, so the compiler then reported a missing allowlist or an empty condition instead of pointing at the layout. Fix. Write the list or map on one line: RESULT_FIELDS: [Status, TransactionIdentifier] and SUCCESS_WHEN: { Status: “Success” }.

W212

Severity: warning Meaning. A SECURE_INPUT PROGRESS_PROMPT block is read as TEXT (the line to speak), AFTER (the wait before it), REPEAT_EVERY (the repeat interval) and REPEAT_TEXT (the line for each repeat). Any other line had nowhere to land and was discarded, so the caller would not hear what the author wrote. Fix. Write the hold line as TEXT: “{{locale.PAYMENT_PROCESSING}}” or TEXT: { en: “Please hold.” }, and the timing as AFTER: 2s and REPEAT_EVERY: 10s.

W213

Severity: warning Meaning. PROGRESS_PROMPT AFTER and REPEAT_EVERY take a whole number followed by s or ms. A value without a unit or with another unit could not be read, so the default was used instead: AFTER waits 0 s on the block and 2 s on a field, and without REPEAT_EVERY the line is spoken once. Fix. Write AFTER: 2s or REPEAT_EVERY: 10s.

W214

Severity: warning Meaning. Branch, ON_START, action handler, SUB_INTENTS, hook, and escalation SET blocks keep one value per name. When a name is assigned twice, the later line replaces the earlier one while the file is read. The earlier line never runs. A FLOW step SET is different: it keeps every line and runs them in written order. Fix. Keep one assignment per name. To build a value over several lines, move the lines to a FLOW step SET, for example in the step this block goes to.

W217

Severity: warning Meaning. A FLOW step header is recognised only as a bare name: line with nothing after the colon, written at column 0 or at two spaces under FLOW:. A header written at another indentation, for example after a comment block, or with a comment after the colon (or trailing spaces at two spaces), was read as an unknown property of the step above it and dropped. The step never existed; its REASONING, ON_RESULT, THEN and other properties were applied to the step above, replacing the properties that step already had, and routes to the missing step could no longer resolve. Nothing reported it at parse, validate or import preview. Fix. Write the header as a bare name: line at the same level as the other FLOW step headers (column 0 or two spaces under FLOW:), with no comment or trailing spaces after the colon.

W218

Severity: warning Meaning. A step has one slot for each of REASONING, GOAL, THEN, ON_INPUT, ON_RESULT, CALL, RESPOND and the other single-value properties. Writing one a second time silently kept only one of the two values, normally the later, so authored logic, including a whole ON_RESULT branch list, could disappear with no message. This often follows a mis-indented step header (W217), whose properties land on the step above. Fix. Keep one value per property in each step, or move the second one into its own step and route to it with THEN.

W324

Severity: warning Meaning. HAND_BACK takes after_reply or when_finished. A handoff with any other value keeps the after_reply behavior. A reasoning specialist with no COMPLETE condition is handed back after its reply. Fix. Use HAND_BACK: after_reply or HAND_BACK: when_finished, or remove the line. See Handing back.

W325

Severity: warning Meaning. HAND_BACK sets when a returnable handoff hands the customer back to this agent. A handoff without EXPECT_RETURN: true is a permanent transfer, so there is nothing to hand back. The line has no effect. Fix. Add EXPECT_RETURN: true to the handoff, or remove HAND_BACK.

W362

Severity: warning Meaning. COMPLETE: holds a list of conditions, and each one starts with - WHEN:. RESPOND, SET, STORE, TERMINAL, and response payloads are properties of one condition. The runtime reads them only when they are indented under that - WHEN: line. Anything written directly under COMPLETE: is ignored. The agent never completes from it, and the goodbye it holds is never spoken. Fix. Start the block with a condition line, such as - WHEN: true, and indent RESPOND, SET, and the other properties under it. Use a condition that is true when the agent is done. You can also end a FLOW step with THEN: COMPLETE.

W363

Severity: warning Meaning. A COMPLETE condition reads RESPOND, MODEL_CONTEXT, SET, STORE, TERMINAL, and response payload blocks. For compatibility, the runtime still reads a property written at or left of the indent of its - WHEN: line into that condition. For example, TERMINAL: false there keeps the session open. This layout looks like content written directly under COMPLETE:, which is otherwise ignored. Fix. Indent the property under its - WHEN: line. If it belongs to a different condition, move it there. If it is not needed, remove it.

W380

Severity: warning Meaning. A DO: list under an ON_START BRANCHES entry uses the same grammar as a FLOW step DO:. An entry that grammar cannot read is reported here as a warning instead of an error, because an ON_START branch DO: was ignored before, and the rest of the list still runs. Fix. Write each entry as a ”- SET:”, ”- CALL:” or ”- HANDOFF:” line, as in a FLOW step DO: list.

W388

Severity: warning Meaning. The top-level on_start do list hands off at most once, to a named agent. An empty handoff, or a second one, is ignored. Fix. Keep one handoff entry that names the target agent.

W408

Severity: warning Meaning. A ROW groups the action items that share one line. A ROW with no BUTTON, SELECT, INPUT, SUBMIT or TEXT item under it places nothing, so it is dropped from the layout. Fix. Indent the items that belong on this line under the ROW, or remove the empty ROW.

W409

Severity: warning Meaning. A layout is a list of lines, and a ROW is one line. A ROW inside a ROW has no meaning, so the nested ROW is skipped together with every item under it. Fix. Move the nested items into the outer ROW, or make them a separate ROW at the ACTIONS level.

W410

Severity: warning Meaning. WIDTH is an item’s share of its ROW relative to the other items on that line: an item with WIDTH 2 takes twice the space of an item with WIDTH 1. Any other value is ignored and the item uses the default share of 1. Fix. Set WIDTH to a whole number from 1 to 12, for example WIDTH: 2.

W411

Severity: warning Meaning. A form has one submit button, so the SUBMIT marker can place it only once. Every SUBMIT after the first is ignored. Fix. Keep one SUBMIT marker where the submit button should appear.

W412

Severity: warning Meaning. The SUBMIT marker only places the submit button. The button itself is declared by SUBMIT_LABEL (its text) and SUBMIT_ID (the action id ON_ACTION handles). Without both there is no button to place, so the marker is removed. Fix. Add SUBMIT_LABEL and SUBMIT_ID to the ACTIONS block, or remove the SUBMIT marker.

W413

Severity: warning Meaning. A TEXT item shows a line of static text inside the form. An empty TEXT shows nothing, so it is dropped from the layout. Fix. Write the text after TEXT:, for example - TEXT: “Need help? Call support.”

W414

Severity: warning Meaning. The SUBMIT marker only decides where the submit button sits. Its text always comes from SUBMIT_LABEL and its action id from SUBMIT_ID, so a value written after SUBMIT: is ignored. The marker itself still places the button. Fix. Write - SUBMIT on its own and put the text in SUBMIT_LABEL, for example SUBMIT_LABEL: “Sign in”.

W494

Severity: warning Meaning. A FEEDBACK block can override the voice settings of the survey template it references: the spoken prompts (voice_prompts: keypad_instruction, retry_invalid, timeout, error), the number of retries after a wrong key (voice_retries, 0 to 3) and the no-input timeout in seconds (voice_timeout, 3 to 30). A value the template rules would refuse is not used, so callers hear the template’s setting, or the platform default when the template has none. Fix. Use a listed prompt name, remove any {{…}} placeholder from the prompt text, or bring the number into its range.

W496

Severity: warning Meaning. On ON_START and HOOKS calls, WHEN: guards whether the tool call runs only when it is indented under the CALL: it belongs to. Written as a sibling of CALL: nothing reads it, so the call runs unconditionally, exactly as it did before the nested guard existed. Fix. Indent WHEN: one level under the CALL: it should guard, or remove it.

W497

Severity: warning Meaning. The opt-in EXECUTION keys accept a fixed set of values. escalation_recheck_after_tools and guardrail_builtins take true or false. set_expression_failure takes keep_text or unset. prompt_cache_key takes true, false, or a name of up to 64 letters, digits, underscores, dots, colons, or hyphens. prompt_cache_retention takes in_memory or 24h. prompt_version takes v1 or v2. The older values off and auto read as v1 and v2. The compiler drops any other value, so the key behaves as if it were absent and uses its default. For set_expression_failure the default is unset, so a misspelled keep_text does not keep the text. Fix. Use one of the values listed above.

W498

Severity: warning Meaning. A step PERSONA: line, or step behavior-profile syntax in a YAML flow step (persona or use_behavior_profile keys) or in legacy STEPS:, is ignored. A step has no persona setting of its own, and these forms do not attach a profile. In a text FLOW step, USE BEHAVIOR_PROFILE: is supported and does not report this code. Fix. Remove the line. In a text FLOW step, attach the profile with USE BEHAVIOR_PROFILE:. Or attach it at agent level and target channels or callers with the profile WHEN. Put step-specific wording in the step GOAL or BEHAVIOR, step rules in STEP_CONSTRAINTS, the step tool list in AVAILABLE_TOOLS, and step voice handling in the step VOICE block. See Where profiles apply.

W562

Severity: warning Meaning. A provider rejects a request that carries more stop sequences than it allows, which would fail the turn. The runtime therefore leaves the whole list out of any call to a provider whose limit it exceeds, and the reply is generated without stop sequences. Fix. Keep stop_sequences to four entries or fewer so every supported provider receives them.

W563

Severity: warning Meaning. Named execution profiles are not supported. A step EXECUTION takes the same keys as the agent EXECUTION section, written one per line under it. The step keeps the agent settings. Fix. Write the settings as indented keys, for example EXECUTION: then model: gpt-4o-mini on the next line.

W564

Severity: warning Meaning. A step EXECUTION value is outside the range the parameter accepts. It is not clamped or coerced. The step uses the agent value for this key and keeps its other step keys. Fix. Set a value that satisfies the constraint quoted in the message, or remove the key.

W565

Severity: warning Meaning. A step EXECUTION value has the wrong type, is not one of the allowed words, or is null. A quoted number is text, not a number. The step uses the agent value for this key and keeps its other step keys. Fix. Write the value in the shape quoted in the message, or remove the key.

W566

Severity: warning Meaning. An EXECUTION header with no keys under it changes nothing for the step. Fix. Add indented keys under EXECUTION:, or remove the header.

W567

Severity: warning Meaning. fallback_model and conversation_history_window apply to the whole agent turn and history, so a step value has no consistent meaning. Fix. Set it in the agent EXECUTION section instead, or remove it from the step.

W568

Severity: warning Meaning. A step EXECUTION key is not recognized. A step block reads the model, the reasoning effort, the output length limit, the sampling settings (temperature, top_p, top_k, the two penalties, and seed), verbosity, the thinking settings, stop_sequences, llm_timeout, and tool_timeout. Fix. Check the spelling against that list, or remove the key.

W569

Severity: warning Meaning. Loop limits, queueing, budgets, compaction, operation models, and voice settings govern the whole agent, not one step. Fix. Set it in the agent EXECUTION section instead, or remove it from the step.

W573

Severity: warning Meaning. Every key in a step EXECUTION block must start at the column of the first key. A line at another column, or one that is not a key: value pair, is skipped with anything nested under it. Fix. Align the key with the other keys in the block.

STEP_EXECUTION_MODEL_CAPABILITY

Severity: warning Meaning. A step EXECUTION names a model the platform’s capability registry knows, and sets a parameter that model does not accept, for example reasoning_effort on a model without reasoning, or temperature on a model that disables it. The runtime removes or adjusts the parameter for that model, so the authored value has no effect. Fix. Remove the parameter from the step, or choose a model that supports it. Models given as {{config.KEY}} templates are not checked at compile time.

STEP_EXECUTION_NON_REASONING_STEP

Severity: warning Meaning. A FLOW step that is not a REASONING step carries an EXECUTION block. Only REASONING steps call a model with step settings, so the block has no effect and is left out of the compiled agent. Fix. Add REASONING: true to the step if it should reason with these settings, or remove the EXECUTION block.

STEP_EXECUTION_STOP_SEQUENCES_WITH_TOOLS

Severity: warning Meaning. A REASONING step sets stop_sequences but offers the model tools. The runtime never sends stop sequences on a call that offers tools, because a match in the text would end generation before the tool call. The list takes effect only on the step’s calls made without tools. Fix. Remove stop_sequences from the step, or give the step AVAILABLE_TOOLS: [] if it should answer without tools.

AGENT_EXECUTION_ROLE_CONFLICT

Severity: error Meaning. The declared BEHAVIOR role contradicts another declaration on the same agent, either the deprecated canRespondDirectly flag or the fact that the document is a SUPERVISOR. Fix. Keep one source of truth: delete the deprecated canRespondDirectly field, or change role to the value it implies. Supervisors must use role “router” or “orchestrator”.

AGENT_RETURNED_WORK_REQUIRES_ORCHESTRATOR

Severity: error Meaning. The agent declares DELEGATE targets but its BEHAVIOR role is not “orchestrator”. Only an orchestrator is allowed to receive delegated work back and synthesize a response from it. Fix. Set BEHAVIOR role to “orchestrator”, or replace DELEGATE with HANDOFF if the other agent should own the conversation from that point on.

CALLER_HOLD_HOOK_RESPOND_UNSUPPORTED

Severity: error Meaning. An on_caller_hold_limit hook has a RESPOND, a WHEN reply, or rich content. The caller is on hold and the call can end right after the hook, so nobody would hear it. Fix. Remove RESPOND, WHEN replies, and rich content from the hook. Use CALL, SET, and HANDOFF only.

CALLER_HOLD_TRANSFER_TARGET_MISSING

Severity: warning Meaning. A project import turned caller hold off. Caller hold transfers the caller to a SIP or phone number target at the hold limit. The project has no valid target of that kind in its Agent Transfer settings. Fix. Add the target in Agent Transfer. Then turn caller hold on.

CALL_WITH_REQUIRED_PARAM_EMPTY

Severity: warning Meaning. A CALL WITH: block binds an empty value ("" or '') to a parameter the tool declares as required. The runtime treats an empty required input as missing. It does not call the tool, and the step takes its failure path with the error code TOOL_REQUIRED_INPUT_MISSING. The warning skips a parameter that has a default, a non-model source, or hidden model visibility. It also skips a parameter that process.before fills, and the session values session_id, tenant_id, and project_id. An empty user_id still warns, because a session with no user has no user_id. Fix. If an empty value is valid, declare the parameter optional in the tool signature (name?: string) so the runtime sends it. Otherwise bind a value that is never empty, or branch on __errorCode == "TOOL_REQUIRED_INPUT_MISSING" in ON_FAILURE.

CEL_CONDITION_UNRESOLVED_CALL

Severity: warning Meaning. A condition parses as CEL but calls a function or method the evaluator has no overload for, such as CONTAINS(x, “y”) written as a bare function call or x[0].has(field). At runtime the call cannot be evaluated, the evaluator falls back and the condition always evaluates false, so the branch, guard, or trigger it controls never fires. It is a warning, not an error, so agents already deployed keep compiling. (MATCHES(…) and MATCH_RESULT(…) are now registered functions, field-format design, and no longer land here.) Fix. Use a supported form. Regex: x.matches(“re”) or x MATCHES “re”. Field presence on a map or list element: “field” in x[0], or has(x.field) on a plain member path. The message names the call the evaluator could not resolve.

CEL_EXPRESSION_LIMIT

Severity: warning Meaning. An expression exceeds the supported CEL budget after preparation. Ordinary boolean conditions allow up to 65536 UTF-16 units, 4096 syntax nodes and 64 nesting levels, with each indivisible expression at most 4096 units. Scalar expressions remain limited to 4096 units. Unsupported expressions retain their existing fallback or refusal policy, and routing may evaluate alternatives separately. This warning does not block compilation or import. Fix. Simplify the named expression or split it into shorter rules. Keep individual comparisons, function calls and scalar SET expressions within 4096 prepared source units. Do not rely on legacy fallback to make an unsupported condition agree across escalation, flow and routing surfaces.

CEL_MATCH_PATTERN_REJECTED

Severity: warning Meaning. A condition or guardrail check passes a literal pattern to abl.matches_pattern or abl.not_matches_pattern (or uses x MATCHES “pattern”, which runs as abl.matches_pattern) that the runtime regex engine cannot run: it is longer than 500 characters, uses a construct RE2 does not support (backreferences, lookahead or lookbehind), or is not a valid regex. At runtime the call does not fail: abl.matches_pattern returns false and abl.not_matches_pattern returns true for every input, so a guardrail built on it never fires (or always fires) and the trace shows an ordinary result. It is a warning, not an error, and evaluation is unchanged. Fix. Keep each pattern to 500 characters or fewer and to RE2 syntax. Split a long alternation into several calls of at most 500 characters each, joined with || for abl.matches_pattern or with && for abl.not_matches_pattern. Replace lookarounds and backreferences with plain alternatives or an additional condition.

CONFIGURATION_INVALID_VALUE

Severity: warning Meaning. A configuration value is not one the setting supports. The parser keeps its existing behavior, and the message says what applies instead. Strict validation does not promote this warning to an error. It is raised for these settings: the EXECUTION conversation_history_window (a whole number from 0 to 10; an invalid value is not applied and the runtime keeps its default of 2), EXECUTION compaction_threshold (a number from 0 to 1), a block-form value for an EXECUTION scalar key, and the HANDOFF CONTEXT history value and its last_n count. Fix. Use a supported value for the named setting. The message names the setting and the behavior that applies instead. If the setting is missing, check the inherited project configuration.

CONFIGURATION_UNKNOWN_KEY

Severity: warning Meaning. A key is not recognized in the block that holds it. Names are scoped to their block, so a key that works in one block can be unknown in another. The key has no effect, and the defaults or inherited settings stay in effect. The warning does not make a misspelled or wrongly cased key work. Strict validation does not promote it to an error. The parser raises it for unknown keys in the agent EXECUTION block (ABL text and YAML) and in a HANDOFF CONTEXT block. In ABL text, EXECUTION keys match without regard to case or underscores. In YAML, only the documented spellings (snake_case or camelCase) are accepted, so a key with other casing warns. Fix. Check the spelling, the case, and the block. See Configuration properties for the supported EXECUTION keys.

CONSTRAINT_ON_FAIL_VOICE_ILLEGAL_PLACEMENT

Severity: warning Meaning. A constraint refusal is spoken with the delivery policy of the step or turn it interrupts. A VOICE, BARGE_IN, STREAMING or LISTENING line there has no effect, so it is reported with a warning instead of being dropped silently. Fix. Remove the line, or set the policy on the FLOW step or the agent CONVERSATION.listening.

CONVERSATION_BACKCHANNEL_DURATION_CAP_IGNORED

Severity: warning Meaning. on_caller_backchannel.max_duration_ms has no effect. The voice gateway cannot enforce a duration cap and reports no utterance duration, so the value is ignored and back-channels are matched on the phrase list alone. Fix. Remove max_duration_ms from on_caller_backchannel.

CONVERSATION_BACKGROUND_DISALLOW_DEFAULT

Severity: warning Meaning. listening.mode is background but listening.barge_in (else listening.on_overlap) resolves to disallow. A background call with barge-in off behaves as a slower turn-taking call. Fix. Set listening.mode to turn_taking, or set listening.barge_in to allow so the caller can interrupt in the background mode.

CONVERSATION_BACKGROUND_ONLY_WITH_TURN_TAKING

Severity: warning Meaning. on_caller_backchannel or barge_in_continuation is authored together with listening.mode turn_taking. Both only work in background listening mode, so they stay inactive on every turn that listens in turn_taking. Fix. Remove the background-only key, or set listening.mode to background (the mode is fixed per call).

DELEGATE_CHILD_GATHER_INPUT_UNMAPPED

Severity: error Meaning. A DELEGATE target requires GATHER fields (required, no default) that the delegate INPUT mapping does not supply. Delegation is a one-shot internal call with no customer turn, so the child would block asking for input nobody can give. Fix. Map every required child gather field in DELEGATE INPUT, give those fields defaults or mark them optional on the child, or model a genuinely interactive child as a HANDOFF with RETURN true so it can converse with the customer before control comes back.

DEPRECATED_INLINE_FEEDBACK

Severity: warning Meaning. An inline FEEDBACK block declares its question and control directly on the agent, with no survey template behind it. It compiles and renders as authored and the customer is still asked. Feedback Insights groups responses by template ID and selects a compatible scoring method at read time; an inline response has no template ID and does not appear in template-scoped score, trend, distribution or segment views. Inline declarations remain supported for compatibility. A project survey template gives agents one reusable, exact-name definition of the question and control without configuring a scoring method on the template. Fix. Create a survey in the project Feedback Templates area and attach it here with FEEDBACK_TEMPLATE, moving the inline prompt, type and labels onto it. The referencing block then keeps only per-string overrides and its render guard, and the responses start appearing in Feedback Insights.

ENABLED_WHEN_ALL_DISABLED_RISK

Severity: warning Meaning. This agent routes with HANDOFF, and every one of its HANDOFF and DELEGATE entries has an ENABLED_WHEN condition, so when all of those conditions are false at once the agent has no route left and must answer every request itself. It is a warning, not an error, so the agent still compiles. Fix. Keep at least one entry without ENABLED_WHEN as an always-available route, or make sure the agent itself can serve the requests that the switched-off routes would have taken.

ENABLED_WHEN_AUTHORED_ROUTE_NO_FALLBACK

Severity: warning Meaning. An authored HANDOFF (in a FLOW step, an action handler, a digression, a tool result or error hook, or an error handler) goes to a target whose every HANDOFF entry has ENABLED_WHEN, and none of those entries has ON_FAILURE. When the condition is false the handoff is refused, and without ON_FAILURE nothing says what the agent does next. It is a warning, not an error, so the agent still compiles. Fix. Add ON_FAILURE (continue, respond or escalate) to the HANDOFF entry for the target, or declare an entry for the target without ENABLED_WHEN if the authored route must always be able to run.

ENABLED_WHEN_EMPTY

Severity: error Meaning. ENABLED_WHEN decides whether a HANDOFF or DELEGATE target is available at all. An empty availability condition is ambiguous: it could mean always available or never available, so the platform refuses to guess and the target is left without one. Fix. Write a deterministic condition such as _metadata.country != “SGP”, or remove ENABLED_WHEN to keep the target always available.

ENABLED_WHEN_LINE_COMMENT

Severity: error Meaning. CEL treats // as a comment to the end of the line. A folded YAML block joins its lines before the condition is read, so a // comment there hides every clause after it and the target could be available when the author meant it not to be. Fix. Use a YAML # comment, or write the condition as a literal block (|) with the comment on its own line.

ENABLED_WHEN_ON_START_UNSTABLE_DATA

Severity: warning Meaning. ON_START delegates to a target whose every DELEGATE entry has ENABLED_WHEN, and a condition reads data that may not exist yet when the session starts, such as conversation memory or tool results. Evaluated at start, such a condition sees empty values, so the delegate may be switched off for reasons the author did not intend. It is a warning, not an error, so the agent still compiles. Fix. Write the ENABLED_WHEN condition over data that is known when the session starts: _metadata, env, caller, session or user. Otherwise, run the delegate later in the conversation instead of in ON_START.

ENABLED_WHEN_SAFETY_ROUTE_CONDITIONAL

Severity: error Meaning. A constraint ON_FAIL hands off to a target whose every HANDOFF entry has ENABLED_WHEN. A constraint handoff is a safety route that must always be able to run, but ENABLED_WHEN can switch the target off, which would leave the failed constraint with nowhere to go. Fix. Declare a HANDOFF entry for the target without ENABLED_WHEN, or send the constraint ON_FAIL to a target that is always available, such as ESCALATE or another unconditional HANDOFF target.

ENABLED_WHEN_SEMANTIC_REFERENCE

Severity: error Meaning. An ENABLED_WHEN condition reads what the current turn means: the intent classification (intent or a field under it), the routing decision (routing.*, routing_intent, source_route_intent), the current user input, or previous_system_message_was_offer, or is natural language that only a model can judge. ENABLED_WHEN decides whether a HANDOFF or DELEGATE target is available at all, before any turn is understood, so it must depend only on data that is known up front. Turn meaning is already expressed by WHEN. Fix. Move the turn-dependent part of the condition into WHEN, and keep ENABLED_WHEN for deterministic availability facts such as _metadata.country != “SGP”, env, caller, session or user data.

ENABLED_WHEN_SHADOWED_BY_UNCONDITIONAL_ENTRY

Severity: warning Meaning. The same target is declared more than once in one HANDOFF or DELEGATE section, and only some of those entries have ENABLED_WHEN. The entry without ENABLED_WHEN keeps the target available whatever the condition says, so the condition does not switch the target off. It is a warning, not an error, so the agent still compiles. Fix. Give every entry for the target the same ENABLED_WHEN condition if the target should be switched off, or remove ENABLED_WHEN from the entry that has it if the target should stay available.

FLOW_ON_INPUT_GATE_AFTER_CONSUMED_INPUT

Severity: warning Meaning. A route already used the customer message (an ON_INPUT branch, or the THEN or ON_RESULT branch of a GATHER step). It reaches a FLOW step that routes only with ON_INPUT and shows no prompt, with nothing shown on the way. Each customer message is routed by one step, so that step does not evaluate the same message. The turn ends with no reply, and the next message is routed there. turn_input does not change this, because the ON_INPUT branches are not evaluated in that turn. Fix. To route on the same message, change the step to ON_RESULT and read the message with turn_input. To wait for a new answer, add a RESPOND with the question on the branch or on a step before it.

FLOW_PROMPTLESS_ON_INPUT_UNMATCHED_TRAP

Severity: warning Meaning. A promptless ON_INPUT-only FLOW step has conditional branches but no ELSE/default branch. Input that matches no condition leaves the session waiting on the same step without an authored prompt. Fix. Add an ELSE branch for unmatched input, add a RESPOND prompt that explains the accepted choices, or keep the gate only when silent waiting is explicitly intended.

FLOW_STEP_WHEN_IGNORED

Severity: warning Meaning. A FLOW step declares a step-level WHEN. The condition is not compiled and has no effect: the step, including its CALL, runs whenever the flow reaches it. Fix. Guard the step from the step before it with ON_RESULT branches, or use CHECK with ON_FAIL to skip it, and remove the WHEN.

GATHER_ASR_REASONING_UNSUPPORTED

Severity: warning Meaning. Listening settings (ASR:) are on a top-level GATHER field of a reasoning agent that is not marked armable: true. The reasoning agent applies a field’s settings only to an armable field, so these are kept for round trips but have no effect on calls. In an agent with a FLOW, a field that sets asr_timeout_ms does not warn: a FLOW tool call missing that input asks for it through this field and applies its settings. Fix. Add armable: true to the field so the agent can apply its listening settings when it asks for it, or remove the ASR: block.

GATHER_ASR_SHARED_STEP

Severity: warning Meaning. A GATHER field with listening settings is asked together with other required fields in the same step. Listening settings apply only when a field is asked on its own, so they never take effect here. Fix. Move the field with ASR: to its own FLOW step.

GATHER_ASR_TERMS_BINDING_FORBIDDEN

Severity: error Meaning. A GATHER field ASR: block takes its special words from a session value the platform never sends to a speech engine: session identity or runtime-internal roots, conversation history, protected email aliases, prototype or secret path parts, or a user.* value without restore_pii: true. Fix. Bind to a value your flow or tools set (for example a list a tool returned). For user.* values, add restore_pii: true only if sending the caller personal data to the speech engine is intended.

GATHER_ASR_TERMS_BINDING_INVALID

Severity: error Meaning. A GATHER field ASR: block takes its special words from a session value, but the value is not exactly one quoted {{path}} reference (surrounding text, several references, filters, an unquoted reference, a reference inside a list, or a path longer than 8 parts or 128 characters). Fix. Write the whole list, or one language list, as a single quoted reference such as terms: “{{caller.streets}}”, where the path is plain dotted names with optional [n] indexes.

GATHER_ASR_UNKNOWN_KEY

Severity: error Meaning. A GATHER field ASR: block names a listening setting the platform does not know. Credential-like keys (service instance ids, API keys, labels) are rejected this way, because listening settings never carry credentials. Fix. Use only provider, model, language, vocabulary (mode, terms) and no_input_timeout_ms, and pick the speech engine by provider name; credentials stay on the voice channel.

GATHER_DTMF_FINISH_KEY_ALONE_INPUT_UNUSED

Severity: warning Meaning. The agent or a behavior profile sets finishKeyAlone to input. A FLOW question it reaches has no finishKeyAlone of its own, and its step has no ON_INPUT branch. A caller who presses only the finish key there hears the question again. Fix. Set finishKeyAlone: value on the question, add an ON_INPUT branch for the key, or set finishKeyAlone: no_input on the question.

GATHER_DTMF_FINISH_KEY_ALONE_INVALID

Severity: error Meaning. finishKeyAlone in keypad settings has a value its level does not allow. It sets what happens when the caller presses only the finish key: no_input (wait and ask again, the default), input (send the key to the agent), or, on a FLOW question only, value (store the key as the answer). Fix. Use no_input or input on a behavior profile, an agent, or the project. Use no_input, input, or value on a FLOW GATHER field.

GATHER_DTMF_FINISH_KEY_ALONE_NEVER_ACCEPTED

Severity: warning Meaning. finishKeyAlone is value on a FLOW question whose allowed values or regular-expression validation reject the finish key. Storing the key always fails validation, and the question is asked again. Fix. Add the finish key to the allowed values or the pattern. Or use input and handle the key in an ON_INPUT branch.

GATHER_DTMF_FINISH_KEY_ALONE_NO_FINISH_KEY

Severity: warning Meaning. finishKeyAlone is input or value in keypad settings whose finishOnKey is none. With no finish key, the caller cannot press it alone, so the setting never applies. Fix. Set finishOnKey to "#" or "*" in the same keypad settings, or remove finishKeyAlone.

GATHER_DTMF_FINISH_KEY_ALONE_VALUE_REASONING

Severity: error Meaning. finishKeyAlone is value on a top-level GATHER field. The reasoning agent collects those fields as a whole, so there is no single question to store the finish key in. Fix. Use input on this field so the agent receives the key. Or set value on the field in the FLOW step that asks for it.

GATHER_DTMF_FINISH_KEY_ALONE_VALUE_TYPE

Severity: error Meaning. finishKeyAlone is value on a FLOW question whose type can never hold a single # or *. These types are number, date, phone number, email address, yes or no answer, list, and range. Fix. Use input and handle the key in an ON_INPUT branch. Or ask for the key in a text or menu question.

GATHER_DTMF_FINISH_KEY_INVALID

Severity: error Meaning. finishOnKey in a FLOW GATHER field DTMF: block is not "#", "*", or "none". It cannot finish the typed answer. Fix. Use "#", "*", or "none".

GENESYS_SIP_HEADER_KEY_COLLISION

Severity: error Meaning. Two or more of the Genesys SIP transfer headers, reason, queue, metadata, resolve to the same header key once a trigger-level GENESYS_SIP_HEADERS override is merged over the ESCALATE routing defaults (the comparison is case-insensitive, as SIP header names are). The runtime payload builder rejects a transfer whose header keys collide, so the escalation would fail during a live call; the compiler reports the collision at authoring time instead. Fix. Give each emitted header a distinct key via reason_header_key, queue_header_key and metadata_header_key. Remember the defaults count as taken, the reason header defaults to X-Transfer-Reason and the metadata header (sent only when a metadata value exists) to X-Transfer-Metadata, and check both the routing block and every trigger override, since the effective set is their merge.

GUARDRAIL_BUILTIN_NOT_BOUND

Severity: warning Meaning. A guardrail check reads one of the identifiers content, session_state or tool_results, or a BEFORE tool-call constraint (REQUIRE or WHEN) reads tool_args, but the agent does not set EXECUTION guardrail_builtins: true and declares no session variable of that name. The name is then not bound, so the check errors when it runs: a guardrail check lets every message through, while a BEFORE tool-call constraint that cannot be evaluated fails closed and blocks the call. It is a warning, not an error, and evaluation is unchanged. Fix. Set guardrail_builtins: true under EXECUTION to bind content (the checked text), session_state, tool_results and tool_args, or write text instead of content. Setting the key changes behaviour: a check that failed open starts to apply and can block.

GUARDRAIL_CONTENT_SHADOWS_SESSION_VAR

Severity: warning Meaning. The agent sets EXECUTION guardrail_builtins: true and also declares a session variable named content, and a guardrail check reads the identifier content. With the alias on, content is the text being checked (an alias of text), so the check tests that text and never the session variable, whatever the author intended. It is a warning, not an error, and evaluation is unchanged. Fix. Rename the session variable so the two no longer share a name, or, if the checked text was meant, write text instead of content to make that explicit.

GUARDRAIL_UNKNOWN_PROPERTY

Severity: warning Meaning. The parser discards a guardrail property it does not recognize. The guardrail then runs with the default for what the property was meant to set. A misspelled action leaves the guardrail on its default action (warn in ABL, block in YAML). A misspelled check leaves it with no check. Fix. Correct the property name. Use the suggestion in the message when there is one. Delete the line if you do not need it. Property names are case-insensitive in ABL and exact in YAML.

HOOK_LANGUAGE_ACTIONS_REQUIRE_SETTING

Severity: warning Meaning. HOOKS before_turn interaction-language actions are authored. They take effect only when the project’s language detection setting is on, on chat and on pipeline voice calls; runtime skips them on realtime speech-to-speech calls. The compiler cannot see project settings, so it reports this on every compile of such an agent. Fix. Turn on language detection in the project settings, or remove the actions if the agent should keep one language.

HOOK_LANGUAGE_ACTION_NOT_BEFORE_TURN

Severity: error Meaning. Runtime evaluates hook interaction-language actions once per turn, before the filler and the model call, so the reply, the filler and the voice all use the language the hook chose. Only before_turn runs at that point; in any other hook the action would have nothing to steer, so it is rejected instead of being silently ignored. Fix. Move the SET_INTERACTION_LANGUAGE or CLEAR_INTERACTION_LANGUAGE action, with its WHEN:, under HOOKS before_turn.

HOOK_LANGUAGE_ACTION_UNSUPPORTED_HOOK

Severity: error Meaning. The IR carries interaction_language_actions on a hook other than before_turn. Runtime runs these actions only in the before_turn early pass, so on any other hook they would be silently ignored. Fix. Move the interaction-language actions to HOOKS before_turn.

HOOK_LANGUAGE_CATCH_ALL_NOT_LAST

Severity: warning Meaning. before_turn interaction-language actions are tried in order and the first whose WHEN: holds wins. An action with no WHEN: always matches, so every action written after it is unreachable. Fix. Move the action without WHEN: to the end of the before_turn language actions, or give it a WHEN: condition.

HOOK_LANGUAGE_ENTRY_INVALID

Severity: error Meaning. Each interaction_language_actions entry is one guarded action: it either sets the session language (from a tag or an expression) or clears it. An entry with none of the keys does nothing, and an entry with several is ambiguous, so the entry is dropped and the agent does not compile. Fix. Keep exactly one of set_interaction_language, set_interaction_language_expression or clear_interaction_language in the entry, next to its optional when.

HOOK_LANGUAGE_WHEN_INVALID_CEL

Severity: warning Meaning. A WHEN: that guards a HOOKS before_turn SET_INTERACTION_LANGUAGE or CLEAR_INTERACTION_LANGUAGE is not a valid CEL expression. The action is dropped rather than run unconditionally, because an always-on language switch would change the language of every reply, filler and voice prompt. Fix. Rewrite the WHEN: as one valid CEL expression, for example session.languageDetection.status == “detected”, or delete the action.

HOOK_VOICE_NOT_UNDER_RESPOND

Severity: warning Meaning. A hook VOICE block styles the reply of the RESPOND it is indented under, including a RESPOND inside a WHEN: branch. Written beside a one-line RESPOND, or in a hook with no RESPOND, nothing reads it, so its BARGE_IN, STREAMING and acoustic settings do not apply and the hook reply keeps the step and turn defaults. After a | block RESPOND in a WHEN: branch the block would swallow a nested VOICE as reply text, so VOICE goes level with that RESPOND instead. Fix. Indent VOICE: one level under the one-line RESPOND: it belongs to (level with the RESPOND: after a | block in a WHEN: branch), or remove it.

HOOK_WHEN_BODY_MULTIPLE_ACTIONS

Severity: error Meaning. A hook WHEN: branch is either a reply branch (RESPOND) or a language branch (SET_INTERACTION_LANGUAGE or CLEAR_INTERACTION_LANGUAGE). The two run at different points of the turn, so one body cannot hold both, and two language actions under one WHEN: would leave it unclear which one applies. The whole branch is dropped. Fix. Split the body into separate WHEN: branches with the same condition: one with the RESPOND and one with the language action.

HOOK_WHEN_BRANCH_INVALID_CEL

Severity: warning Meaning. A hook WHEN: line that precedes an indented RESPOND: is not a valid CEL expression. The branch is not compiled as a conditional branch: the WHEN: is ignored and the RESPOND: runs unconditionally on every run of the hook, exactly as it did before hook branches existed. It is a warning, not an error, so the agent still compiles. Fix. Rewrite the WHEN: as one valid CEL expression so the RESPOND: applies only when it is true, or delete the WHEN: line if the RESPOND: is meant to run every time.

HTTP_TOOL_INVALID_IDEMPOTENCY_HEADER

Severity: warning Meaning. An HTTP tool declares idempotency_header with an empty, malformed, overlong, or reserved header name (a credential, cookie, transport, SOAP or request-signing header). The runtime therefore does not send its idempotency key to the provider for that tool. Fix. Set idempotency_header to the header your provider documents for idempotency keys, for example Idempotency-Key, or remove it if the provider does not support idempotency keys.

IMMUTABLE_SYSTEM_IDENTIFIER

Severity: error Meaning. A SET, CLEAR, or return mapping writes to a system-owned identifier, user_id, session_id, tenant_id, or project_id, including any nested path under one of them. Those values carry the identity and isolation context the platform itself owns, so authored writes to them are refused rather than allowed to corrupt who or where the session is. Fix. Write the value under a domain-specific name instead, customer_user_id, order_session_ref, and map it explicitly wherever it is consumed. Identity itself only moves through the supported channels: handoff pass-through, memory grants, or runtime bootstrap; remove the SET, CLEAR, or map entry that targets the reserved name.

INVALID_APPROVED_SIP_AUTHORITY

Severity: error Meaning. A call_transfer tool declares an approved_authorities allow-list entry that is not a valid literal SIP authority, malformed scheme, missing/non-literal host or port, a query/header section, or another structural defect the shared compileApprovedAuthority parser rejects. Fix. Author the entry as a literal ‘sip:’ or ‘sips:’ URI with an explicit VDN (userinfo), host, and port and no query/header section, for example, ‘sip:421001@172.0.41.7:5060’.

INVALID_CEL_SYNTAX

Severity: error Meaning. A condition that reads as a CEL expression fails to parse, even after legacy-operator migration and presence-check normalization. The runtime evaluator never throws, an unparseable condition silently evaluates false, so the branch, constraint, trigger, or completion it guards would simply never fire, with no error at runtime to say why. Fix. Rewrite the condition as a valid CEL expression: use == and != for equality (never the JavaScript triple forms), && / || / ! or AND / OR / NOT for logic, and keep every parenthesis, bracket, and quote balanced. String values go in double quotes. Test membership with in, strings with contains, matches, startsWith, or endsWith, and presence with has() on a member path or the IS SET / IS NOT SET sugar. Platform helpers live under the abl namespace, for example abl.upper(name) or abl.round(price, 2). For a condition that should always hold, write true. The message carries the parser description of exactly what it could not read.

INVALID_HANDOFF_TARGET

Severity: error Meaning. A handoff target does not resolve: the named agent does not exist in this compilation, or the reference, from ON_ACTION, a digression or an error handler, names an agent this agent holds no handoff authority for. The transfer would have nowhere to go at runtime. Fix. Correct the target to the exact name of an agent in the project, and declare it in this agent HANDOFF configuration (for error handlers, ROUTING or a constraint handoff also grants authority). Names match exactly; a target outside this project needs a declared remote handoff entry.

INVALID_SIP_TRANSFER_DESTINATION

Severity: error Meaning. A call_transfer tool declares a static sipTransferId default that is empty, templated, or not a valid sip: or sips: URI, so the transfer destination cannot be used reliably. Fix. Set default_source to config or system and provide a literal sip: or sips: URI such as ‘sip:desk@sbc.example.com’. Remove the default to use the project-configured transfer target, or configure a dynamic target in Agent Transfer settings instead of templating the value.

INVALID_SURVEY_CONDITION

Severity: error Meaning. A survey template’s follow-up or comment-box condition cannot be evaluated as written: it does not parse, it references a name the platform does not supply, or a matching condition would select a question that does not exist. An invalid condition never matches, so left unresolved the survey ships looking configured and behaves as though it had no conditions at all, a follow-up nobody is ever asked, or a comment box nobody is ever offered. Fix. Fix the condition on the survey template in the project Surveys area, the message names the editor (comment box or follow-up) and the position of the condition at fault. A condition may reference only rating, max, language, channel and session, with session values addressed as session.name. If the condition was meant to select a per-condition question, give it one, or give the follow-up a default question for it to fall back to.

LISTENING_FINISH_KEY_ALONE_VALUE_LEVEL

Severity: error Meaning. finishKeyAlone is value on a behavior profile, an agent, or the project. Storing the finish key as an answer needs a single question to hold it, so value works only on a FLOW GATHER field. Fix. Use input here so the agent receives the key. Or set value on the field in the FLOW step that asks for it.

LISTENING_KEY_NEEDS_NEWER_RUNTIME

Severity: warning Meaning. A listening setting this runtime knows but does not apply yet (for example asr_timeout_ms) was written. It is checked for range and then ignored, so the agent keeps working on older and newer runtimes. Fix. Keep it if the agent will run on a runtime that supports it; otherwise remove it. Nothing changes on this runtime either way.

MATCHES_FUNCTION_REGEX_LITERAL

Severity: warning Meaning. MATCHES(subject, “pattern”) is the fuzzy spoken-answer function that compares what a caller said with a stored value, but its second argument is a string literal that looks like a regex (an anchor, character class or quantifier). The function does not run a regex, so the comparison cannot behave the way the pattern suggests and the condition is almost certainly wrong. It is a warning, not an error, and evaluation is unchanged. Fix. If a regex was meant, write the infix form subject MATCHES “pattern” or the method form subject.matches(“pattern”). If a fuzzy comparison was meant, pass the stored value as the second argument, MATCHES(said, stored, “kind”), or use a literal without regex syntax.

MATCHES_INFIX_LITERAL_AMBIGUOUS

Severity: warning Meaning. A condition uses the regex infix form x MATCHES “literal” with a pattern that has no regex metacharacter, so it reads like the fuzzy spoken-answer function MATCHES(a, b). The infix form is the one that applies: it is true when that exact, case-sensitive text appears anywhere in x. It is a warning, not an error, and evaluation is unchanged. Fix. If a fuzzy comparison of what the caller said was meant, write the function form MATCHES(x, “literal”) (or compare two fields: MATCHES(supplied, stored)). If a regex was meant, keep the infix form; anchoring it (^literal$) or using x == “literal” for an exact comparison also clears the warning.

MEMORY_INVALID_ACCESS

Severity: warning Meaning. ACCESS sets whether the agent may read, write, or both read and write a persistent memory value. A value outside read, write, and readwrite restricts nothing. In ABL the value is ignored and the default readwrite applies. In YAML, where values are case-sensitive, the value is kept but acts as readwrite. A value meant to be read-only can be overwritten. Fix. Set ACCESS to read, write, or readwrite (lowercase in YAML). Or remove the line to keep the default.

MEMORY_INVALID_SCOPE

Severity: warning Meaning. SCOPE sets who shares a persistent memory value. user keeps one value per end user. project shares one value across the project. execution_tree shares it among the agents of one handoff tree. A value outside these never selects the scope you meant. In ABL the value is ignored and the path falls back to user, so a value meant to be shared is stored separately for each user. In YAML, where values are case-sensitive, the value matches no scope. The agent then does not load or save the path with its other persistent memory, and tool reads and writes treat it as user. Fix. Set SCOPE to user, project, or execution_tree (lowercase in YAML). Or remove the line to keep the default.

MEMORY_MODEL_VISIBLE_PATH_UNSUPPORTED

Severity: error Meaning. Automatic memory projection must be able to remove the exact declared path. Unsupported selectors cannot safely enforce an authored opt-out. Fix. Use numeric dot segments for array positions, or opt out the supported parent path instead. Existing entries without a false opt-out are unaffected.

MEMORY_MODEL_WRITERS_INVALID

Severity: error Meaning. Explicit model ownership must be a consistent list of agent names. Silently dropping malformed ownership could reopen a protected model write. Fix. Use MODEL_WRITERS: [agent_name, …], or MODEL_WRITERS: [] to deny all model writers. Repeated declarations must agree.

MODEL_CONTEXT_INVALID

Severity: error Meaning. Tool result MODEL_CONTEXT selects either FIELDS or TEMPLATE. Response MODEL_CONTEXT must be a text scalar beside RESPOND. Invalid input cannot safely define a model view. Fix. Use exactly one non-empty FIELDS: [path, …] or TEMPLATE: “text” for tools; use a text scalar beside FLOW, COMPLETE, or tool-hook RESPOND.

NO_INPUT_HOOK_RESPOND_UNSUPPORTED

Severity: error Meaning. An on_no_input_exhausted hook has a RESPOND. The call can end straight after the hook, which could cut the reply off. Fix. Remove RESPOND and put the final words in exhausted_message.

NO_INPUT_TARGET_NO_HOOK

Severity: error Meaning. on_exhausted is hook, but the agent has no on_no_input_exhausted hook, so nothing would run when silence is exhausted. Fix. Add HOOKS on_no_input_exhausted to this agent, or set on_exhausted to hangup or custom.

ON_ERROR_HANDLER_TYPE_DRIFT

Severity: warning Meaning. The ON_ERROR handler type is not one of the canonical runtime error types (default, llm_error, tool_error, routing_error, validation_error, timeout_error, unknown_error). Runtime dispatches on the canonical names, so a handler declared under any other name may never match. Names with a known replacement get a targeted hint; the rest get a generic one. Fix. Rename the handler to a canonical error type, and express the narrower distinction with SUBTYPE, for example tool_error with SUBTYPE: tool_timeout.

ON_START_BRANCH_DO_ACTIVATED

Severity: warning Meaning. Notice: this ON_START branch runs its DO: list (SET, CALL and HANDOFF) at session start when it matches. Before ON_START DO: lists were supported, a branch DO: was silently ignored, so a project that already contained one now runs it. Fix. No action is needed if the actions are intended. Otherwise remove the DO: list from the branch.

ON_START_DELEGATE_WITH_HANDOFF

Severity: warning Meaning. ON_START both delegates and hands off. When a handoff runs at session start, the child owns the session and the ON_START DELEGATE is not returned. Fix. Choose one start-time route: keep the handoff and remove DELEGATE, or remove the handoff and keep the delegation.

ON_START_HANDOFF_ACTIVATED

Severity: warning Meaning. Notice: this agent hands off at session start when no ON_START branch matches. Before ON_START handoffs were supported, this HANDOFF line was silently dropped, so a project that already contained it now behaves differently. Fix. No action is needed if the start-time handoff is intended. Otherwise remove the HANDOFF line from ON_START.

ON_START_HANDOFF_TARGET_INVALID

Severity: warning Meaning. An ON_START handoff names a target this agent cannot start at session start: the agent itself, an agent not in this compilation, an agent with no top-level HANDOFF entry, or a remote (A2A) entry. The entry stays in the compiled agent; at runtime the existing route authority refuses it, the entry agent keeps the session and the refusal is traced. Fix. Name a local agent of this project that is declared in this agent top-level HANDOFF section, and set RETURN, CONTEXT and WHEN on that HANDOFF entry.

ON_START_HANDOFF_UNREACHABLE

Severity: warning Meaning. The top-level ON_START HANDOFF runs only when no ON_START branch matches, and an ELSE branch always matches, so the top-level handoff can never run. Fix. Move the HANDOFF into the DO: list of the branches that should hand off (including ELSE), or remove the ELSE branch.

PARALLEL_CROSS_CALL_READ

Severity: error Meaning. An expression in a PARALLEL call reads another call’s result without AFTER. It can also read a value that the block writes only after all calls finish (SET targets, ERROR_AS, STATUS_AS). Calls run at the same time, so the value is not there yet. Fix. Add AFTER: <name> to wait for the other call, or read the value in a later step.

PARALLEL_DEPENDENCY_INVALID

Severity: error Meaning. An AFTER entry names no call in the block, names its own call, forms a cycle, chains deeper than 3, or is used with a WAIT other than all. Fix. Point AFTER at the AS of another call in the block. Remove the cycle. Keep WAIT: all.

PARALLEL_EXPRESSION_INVALID

Severity: error Meaning. A WHEN, FILTER, KEY_BY, or SET expression in a PARALLEL call does not parse. Fix. Fix the expression syntax.

PARALLEL_LIMIT_EXCEEDED

Severity: error Meaning. A PARALLEL block exceeds a platform limit: 10 calls, 20 requests per batch (each FOR_EACH counts its MAX_ITEMS), MAX_ITEMS 20, MAX_CONCURRENCY 1 to 10, RETRY 3, or TIMEOUT 120 seconds. Fix. Lower the value named in the message, or split the work across steps.

PARALLEL_MAPPING_TARGET_OVERLAP

Severity: warning Meaning. Two tools in one PARALLEL block write the same variable in their on_result.set mappings. Results bind in the order the calls are written, so the later call’s value wins. Fix. Give the mappings different targets, or accept that the later call wins.

PARALLEL_NAME_CONFLICT

Severity: error Meaning. Two names in a PARALLEL block collide (AS, ERROR_AS, STATUS_AS, SET targets, item names). The name result is also reserved for the call result. Fix. Give each value its own name.

PARALLEL_OPTION_CONFLICT

Severity: error Meaning. PARALLEL options contradict each other. Examples: REQUIRED or FAIL_FAST with a WAIT other than all, WAIT n larger than the number of calls, a FALLBACK that is the call’s own tool, or a FALLBACK that does not accept the call’s WITH inputs. Fix. Adjust the options as the message describes.

PARALLEL_REQUIRED_WITH_ON_RESULT

Severity: warning Meaning. A PARALLEL step has a REQUIRED call and ON_RESULT branches. A matching ON_RESULT branch runs before ON_FAILURE, so it can route the step even when a required call failed. Fix. Check the status view in the ON_RESULT branch, or route only through ON_SUCCESS, ON_PARTIAL, and ON_FAILURE.

PARALLEL_SOURCE_INVALID

Severity: error Meaning. The PARALLEL block has a parse error. Examples: an unknown key, a value in the wrong form, a FOR_EACH-only key without FOR_EACH, or no calls. The block does not run, because running it without the bad part would do something other than what you wrote. Fix. Fix the key or value that the message names. Durations are 1500ms, 2s, or a whole number of milliseconds. WAIT is all, first, first_success, or a number.

PARALLEL_STEP_SHAPE

Severity: error Meaning. A FLOW step combines PARALLEL with CALL, GATHER, SECURE_INPUT, HUMAN_APPROVAL, or REASONING: true. Or it uses ON_PARTIAL without PARALLEL. A PARALLEL step only runs its calls and routes on their outcome. Fix. Move the other construct to its own step before or after the PARALLEL step. Or add a PARALLEL block for ON_PARTIAL.

PARALLEL_TOOL_NOT_ALLOWED

Severity: error Meaning. A PARALLEL call or FALLBACK uses a tool that needs user confirmation, publishes SearchAI citations, captures or consumes vault secrets, or has lifecycle hooks that do more than SET, CLEAR, or LOG. Such tools change conversation state while they run. A parallel batch cannot keep that state in order. Fix. Call the tool in its own CALL step, or limit its hooks to SET, CLEAR, and LOG.

PARALLEL_TOOL_NOT_READ_ONLY

Severity: error Meaning. A PARALLEL call or FALLBACK uses a tool that may change data. Only read-only tools run in parallel, because several writes at once could commit in part or in the wrong order. A tool with no declared effect counts as a write, unless it is an HTTP GET, HEAD, or OPTIONS tool. Fix. Declare effect: read_only on the tool if it only reads data. Or call the tool in its own CALL step.

PARALLEL_TOOL_UNKNOWN

Severity: error Meaning. A PARALLEL call or FALLBACK names a tool the agent does not have. Fix. Check the tool name, or add the tool to the agent.

PLAINTEXT_HTTP_HEADER_SECRET

Severity: warning Meaning. An HTTP tool header contains a high-confidence credential-like literal. The compiler preserves the authored request for compatibility, but the value can cross export, logging or runtime boundaries without secret-management controls. Fix. Replace the literal with a secret-backed config reference such as {{config.API_TOKEN}}, or attach the tool to an auth profile. Keep credentials out of authored HTTP header values.

PLAIN_ENGLISH_COMPLETE_WHEN

Severity: error Meaning. The WHEN of a COMPLETE condition is prose, multi-word text with none of the structured comparison, logic, or function forms, but COMPLETE WHEN is evaluated deterministically, not read by the model. The sentence would never be true, so the agent could never complete through this condition. Fix. Restate the condition as a boolean expression over session state, for example task_done == true or refund_status == “issued”. If the judgment the sentence describes genuinely belongs to the model, have a reasoning step or instruction set a variable recording the outcome, then complete on that variable, or classify into intent state first and complete on the classified value.

PROFILE_INVALID_APPLIES_TO

Severity: error Meaning. The behavior profile has an APPLIES_TO line with a value other than all_agents. APPLIES_TO only widens a profile to every agent in the project, so no other value is supported. Compile fails for that profile. Fix. Write APPLIES_TO: all_agents to attach the profile to every agent. Or delete the line and reference the profile from each agent with USE BEHAVIOR_PROFILE:.

REALTIME_TOOL_EFFECT_REQUIRED

Severity: error Meaning. Strict realtime effect metadata mode found a business tool without an explicit effect, so the compiler cannot safely select its guarded scheduler lane. Fix. Add effect: read_only, state_write, or side_effect. Keep compatibility mode only while migrating legacy tools; missing metadata there retains the legacy queue/idempotency path.

RESERVED_VARIABLE_NAME

Severity: warning Meaning. A SET assignment writes to a system-reserved variable name. The runtime uses those names for its own values, regex capture groups among them, and may overwrite the assignment mid-turn. Fix. Rename the variable to one outside the reserved set; a short agent-specific prefix is enough to keep it clear of runtime names.

ROUTE_WHEN_MULTILINE_JOINED

Severity: warning Meaning. A HANDOFF or DELEGATE condition written on indented lines below an empty WHEN: or ENABLED_WHEN:, or continued below an inline ENABLED_WHEN: value, is joined with single spaces into one condition. A line that looks like a key or starts a new entry ends the condition, and # lines are skipped. A condition opened with an explicit block indicator (| or >) is joined the same way without this warning. Fix. Check that the joined condition is the one you meant. Write it on one line, or open it with a block indicator (ENABLED_WHEN: |), to silence this warning.

SECRET_REF_UNSUPPORTED_CONSUMER_TYPE

Severity: error Meaning. A tool declares SECRET_ACCESS read of a key, but its type cannot consume secrets. Only http tools resolve {{_secret.<key>}}; every other type routes to an executor with no secret-substitution step, and a tool with no type at all falls through to the contract-only fallback. The declaration is not merely inert: the runtime takes the value from the single-use vault on the way in, so the entry is spent without ever being delivered, and the next legitimately authorized consumer fails with “already used”, a failure that points at the wrong tool. RESULT_SECRETS capture is unaffected and still works on any tool type. Fix. Move the call that needs the secret into an http tool, or remove the SECRET_ACCESS block from this one. If a workflow was chosen because a fixed body template could not build a variable-length batch payload, an http tool can now build one directly with {{#each input.<parameter>}}, keeping the secret inside the only sink that resolves it.

SECURE_EMBEDDED_PLACEHOLDER

Severity: error Meaning. A SECURE_INPUT deliver.args value embeds a {{secure.<field>}} placeholder inside a larger string (for example, “pan-{{secure.card_number}}”) instead of using it as the whole argument value. Secure field values never pass through the LLM or session state, concatenating them into other text would force the runtime to materialize the raw PCI-scoped value outside the secure delivery path. Fix. Deliver each secure field as its own whole-value placeholder, for example, { pan: “{{secure.card_number}}” }, never interpolated into a longer string. Non-secure context can still use {{ctx.*}} placeholders freely.

SECURE_INPUT_EMPTY_OUTCOME_ROUTE

Severity: error Meaning. A SECURE_INPUT outcome route (ON_SUCCESS, ON_TIMEOUT, ON_CANCEL or ON_FAIL) is written but names no step. These routes take a FLOW step name only; a RESPOND or other block indented under one is not part of SECURE_INPUT and is ignored. With the route empty, the flow has nowhere to go after that outcome, so after a real payment the caller hears the empty-response fallback instead of a confirmation. Fix. Name a FLOW step in the route, for example, ON_SUCCESS: payment_done, and write the RESPOND (and any other work) in that step.

SECURE_INPUT_FIELD_VOICE_ON_SECRET

Severity: error Meaning. A SECURE_INPUT field sets MODE, RECORD, or SPEECH_AS but is not marked NOT_SECRET. These settings let a field be spoken, recorded, and shown in the transcript. They are allowed only on a value you declared readable. A secret field (card number, expiry, security code) always uses the step MODE, is never recorded, and never shows its value. Fix. Remove MODE, RECORD, and SPEECH_AS from the secret field. For a value such as a postal code or a consent answer, add NOT_SECRET: true and SAVE_AS instead.

SECURE_INPUT_INVALID_FIELD_MODE

Severity: error Meaning. A NOT_SECRET field of a SECURE_INPUT step sets a MODE that is empty or names a channel other than dtmf or speech. A mistyped word such as spech would be ignored, and the field would have fewer ways to be answered than you wrote, or none. Fix. Set the field MODE to a non-empty list of dtmf and speech, for example MODE: [dtmf, speech].

SECURE_INPUT_INVALID_MODE

Severity: error Meaning. A SECURE_INPUT step declares a MODE that is empty or includes a collection channel other than dtmf or speech. The DSL parser silently drops any unrecognized mode token, so an author who mistypes a mode ends up with an empty array and a gate the runtime cannot open. Fix. Set MODE to a non-empty list drawn only from dtmf and speech, for example, MODE: [dtmf] or MODE: [dtmf, speech].

SECURE_INPUT_INVALID_NOT_SECRET

Severity: error Meaning. A SECURE_INPUT field marked NOT_SECRET breaks a rule that keeps card data secret. NOT_SECRET takes an allow-listed kind (postal_code) together with SAVE_AS, never READ_ONCE, a luhn or date_mmYY check, or a field named like card data or a PIN, and 5 to 10 digits. Once accepted the value is a session variable, so later CALLs read that variable, not {{secure.<field>}}. The step must keep at least one secret field, and the delivery must send one. Fix. Write NOT_SECRET: postal_code with SAVE_AS: billing_zip and COLLECT: { digits: 5..5 }, and pass “{{billing_zip}}” to the CALL. Leave card numbers, expiry dates, security codes and PINs secret.

SECURE_INPUT_INVALID_PROGRESS_PROMPT

Severity: error Meaning. A SECURE_INPUT PROGRESS_PROMPT has no TEXT, or its timing is out of range. The hold line needs words to speak; AFTER may be at most 60 seconds and REPEAT_EVERY must be 5 to 60 seconds, so a repeat does not cut off the line before it or leave the caller waiting too long. Fix. Write TEXT: “{{locale.PAYMENT_PROCESSING}}” (or a language map), AFTER: 0s to 60s and REPEAT_EVERY: 5s to 60s.

SECURE_INPUT_INVALID_RESULT_CONTRACT

Severity: error Meaning. The SECURE_INPUT delivery CALL result contract cannot be honoured exactly: AS is not a variable name or has no RESULT_FIELDS allowlist, RESULT_FIELDS has no AS to bind to or lists an invalid path, a duplicate or a top-level path named failure (reserved: on ON_FAIL the step writes failure.reason and failure.http_status under AS), or SUCCESS_WHEN is empty, names an invalid path, or lists no accepted value. A path is a response field name or a dotted path of at most five names (result.status). Nothing from a secure response leaves the secure boundary except the scalar fields RESULT_FIELDS names, and SUCCESS_WHEN is exact equality on the named fields. Fix. Write the contract as AS: <variable>, RESULT_FIELDS: [Field, result.status, …] with response field names or dotted paths of at most five names, and optionally SUCCESS_WHEN: { Field: “value” } or { result.status: [value1, value2] }.

SECURE_INPUT_INVALID_SAVE_AS

Severity: error Meaning. A SECURE_INPUT field SAVE_AS is not a variable name, or the step already writes that name as the delivery AS, SAFE_DETAILS_AS or another field’s SAVE_AS. Each would replace the other. Fix. Give SAVE_AS its own variable name of letters, digits and underscores, for example, SAVE_AS: billing_zip.

SECURE_INPUT_INVALID_SPEECH_AS

Severity: error Meaning. A NOT_SECRET field of a SECURE_INPUT step sets SPEECH_AS to a value other than digits or text. digits reads a spoken answer as a number (one one one twenty five is 11125). text keeps the caller’s words. Fix. Write SPEECH_AS: digits for a spoken number, or SPEECH_AS: text for words such as a yes or no answer.

SECURE_INPUT_MISSING_DELIVER_TOOL

Severity: error Meaning. A SECURE_INPUT step has no deliver tool. The parser defaults deliver to an empty tool name when the step has no CALL line, which leaves the collected secure fields with nowhere to go, the gate would collect PCI-scoped values and then have no tool invocation to deliver them to. Fix. Add a CALL <tool> WITH: { … } line inside the SECURE_INPUT block naming the tool that should receive the collected fields.

SECURE_INPUT_MISSING_PROMPT

Severity: error Meaning. A SECURE_INPUT step either declares no FIELDS at all, or declares a field with no usable PROMPT, neither a language map with at least one entry nor a template reference. A field the runtime cannot prompt for can never be collected. Fix. Declare at least one field under FIELDS, and give each field a PROMPT, either a language-keyed map such as { en: “Enter card number” } or a { template: <name> } reference to a resolved survey template.

SECURE_INPUT_NOT_SECRET_CARD_DATA

Severity: warning Meaning. A NOT_SECRET field is checked with luhn or date_mmYY. The card number or expiry is then saved where later steps, tools, and reports can read it. The compiler still accepts it. Fix. Remove NOT_SECRET so the field stays secret, or check a value that is safe to read.

SECURE_INPUT_PROGRESS_PROMPT_UNUSED

Severity: warning Meaning. A SECURE_INPUT field sets PROGRESS_PROMPT, but the field has no VALIDATE CALL. A field hold line is spoken only while that field’s VALIDATE CALL runs; built-in checks such as luhn or regex finish at once and never speak it. Fix. Remove the field PROGRESS_PROMPT, or move it to the SECURE_INPUT block to speak it during the delivery CALL.

SECURE_INPUT_PROMPT_SECURE_REFERENCE

Severity: error Meaning. A SECURE_INPUT field RETRY_PROMPT or NO_INPUT_PROMPT contains a {{secure.<field>}} reference. Prompts are spoken to the caller and go through session state and traces, so a collected card value may never appear in one. Fix. Remove the {{secure.*}} reference from the prompt. Session variables such as {{amount}} and locale entries such as {{locale.CARD_RETRY}} are allowed; for the card, use the safe details from SAFE_DETAILS_AS (for example, {{card.card_number.last4}}) in a later step.

SECURE_INPUT_SPEECH_AS_UNUSED

Severity: warning Meaning. A NOT_SECRET field of a SECURE_INPUT step sets SPEECH_AS, but neither the field MODE nor the step MODE allows speech. The field is answered only on the keypad, so SPEECH_AS has no effect. Fix. Add speech to the field MODE (MODE: [dtmf, speech]), or remove SPEECH_AS.

SECURE_UNKNOWN_FIELD

Severity: error Meaning. A SECURE_INPUT deliver.args value references {{secure.<name>}} for a field name that is not declared under this step’s FIELDS. The runtime has no collected value to substitute, so delivery cannot succeed. Fix. Either add a field with that name under FIELDS, or fix the {{secure.*}} reference in deliver.args to name a field that is actually declared.

SESSION_MODEL_WRITERS_INVALID

Severity: error Meaning. A session MODEL_WRITERS declaration is not a list of non-empty agent names. Fix. Use MODEL_WRITERS: [agent_name, …], or [] to deny all model writers. Omit the declaration to preserve inferred ownership.

SESSION_MODEL_WRITER_UNKNOWN

Severity: error Meaning. A session MODEL_WRITERS declaration names an agent outside the compiled project. Fix. Name a project agent exactly as declared, including case, or remove it from the writer list.

SESSION_READONLY_DECLARATION_MISMATCH

Severity: warning Meaning. Agents disagree on READONLY for the same shared session field. READONLY applies to every agent of the project, so the model of an agent that declares the field writable is refused too. Fix. Declare the field READONLY in every agent that declares it, or remove READONLY where another agent’s model must write it. Authored writes such as SET, on_result, hooks and CONTEXT.pass are not affected.

SET_PLACEHOLDER_IN_EXPRESSION

Severity: warning Meaning. A SET value is a formula (a function call, a comparison, or arithmetic) that also holds a {{name}} placeholder. The runtime never evaluates a SET value that has a placeholder. It fills in the placeholder and stores the whole formula as text. A condition then reads that text as a non-empty value, so a gate such as LOWER("{{flag}}") == "true" is always on. Fix. To compute the value, name the variable without braces: LOWER(flag) == "true". To store text, put the whole value in quotes.

SET_QUOTED_VALUE_READ_AS_PATH

Severity: warning Meaning. In ON_INPUT, ON_RESULT, ON_SUCCESS and ON_FAILURE branches, digression DO blocks, and handoff return handlers, a SET value written in quotes is stored without its quotes. A quoted bare name or dotted path, such as "USER_REQUESTED_FOR_AGENT_TRANSFER", is then read as the variable or path of that name when one exists. It is stored as text only when none exists. A step-level SET keeps the quotes and always stores the text. The warning applies to an agent compiled in legacy SET value mode. A project created with literal SET value mode stores a complete quoted value as written and is not warned. A project created before the mode existed stays in legacy mode. Fix. Move the assignment to a step-level SET when the value must always be text. Or use text that cannot match a variable name, for example text with a space or punctuation. Quoted text with spaces or punctuation is stored as written and does not warn.

STEP_PROFILE_ALREADY_AGENT_LEVEL

Severity: warning Meaning. A FLOW step references a profile that is already attached to the agent. The profile applies at agent scope only, and the step reference changes nothing. Fix. Remove the step reference if the profile should apply across the agent. Remove the agent attachment if it should apply only to this step.

STEP_PROFILE_CALL_SCOPED_FIELD

Severity: warning Meaning. A step-attached profile sets CONVERSATION.listening.mode. The mode is fixed for the whole call, so a step ignores it. Fix. Set CONVERSATION.listening.mode on the agent or in the channel configuration.

STEP_PROFILE_NOT_FOUND

Severity: warning Meaning. A FLOW step references a behavior profile that is not in the compile set. The compiler leaves that step reference out. CI/CD release builds do not load profile documents, so step references on those paths report this code. Fix. Add the profile to the project, or correct the step reference to match the profile name exactly.

STEP_PROFILE_PRIORITY_CONFLICT

Severity: warning Meaning. Two different behavior profiles attached to the same FLOW step declare the same PRIORITY. Their merge order then depends on declaration order. Fix. Give the step-attached profiles different PRIORITY values. Repeated references to the same profile are removed automatically.

STEP_PROFILE_UNKNOWN_TOOL

Severity: warning Meaning. A step-attached profile hides a tool that the agent does not declare, so the hide has no effect. Fix. Correct the TOOLS.HIDE name to a tool the agent declares, declare the tool, or remove the entry.

STEP_PROFILE_UNSUPPORTED_FIELD

Severity: warning Meaning. A step-attached profile contains RESPONSE_RULES, GATHER_OVERRIDES, FLOW modifications, or FLOW_REPLACE. These are not applied at step scope. Fix. Move the field to an agent-attached profile if it should affect the agent, or remove it from the step profile.

STEP_SET_DUPLICATE_TARGET

Severity: warning Meaning. A FLOW step SET block writes the same variable on more than one line. Once you publish the agent again, the lines run one by one in written order. Each line reads the value the line before it wrote: x = 1 then x = x + 1 gives 2. A version published before keeps only the last line for that variable, run where the first line is, so the earlier lines never run there. Fix. Check that the result in written order is the one you want. Then publish the agent so live sessions use it. To keep the old result, delete the earlier lines for that variable.

SYSTEM_PROMPT_LIMITATIONS_APPENDED

Severity: warning Meaning. The agent declares LIMITATIONS and its own SYSTEM_PROMPT, but the prompt has no {{limitations}} placeholder. The runtime adds the limitations after the prompt text under a “Limitations:” heading, so the model still receives every rule; this warning tells you where they end up. Fix. Put {{limitations}} in SYSTEM_PROMPT where the rules should appear (it renders as one ”- rule” line per limitation), or keep the default placement after the prompt and ignore this warning.

TOOL_CACHE_INELIGIBLE

Severity: warning Meaning. The runtime reuses a tool result within one conversation only for read-only tools. An HTTP tool is read-only when its method is GET or HEAD, or when it declares effect: read_only. MCP, lambda, and synchronous workflow tools must declare effect: read_only. Connector, platform, and async webhook tools, async workflows, and system tools are never cached. The hint stays, and the tool runs on every call. Fix. Remove cacheable: true, or declare effect: read_only if the tool never changes data. Strict validation reports this as an error for a tool that declares a write effect.

TOOL_CACHE_INVALID

Severity: warning Meaning. cache_ttl_ms must be a whole number of milliseconds from 1000 to 3600000. The runtime clamps a value outside the range and replaces any other value with the default of 60000. cache_invalidated_by must list other tools of the project that can change data. A tool cannot list itself. Both properties take effect only with cacheable: true. Fix. Set cache_ttl_ms to a whole number from 1000 to 3600000. List only write tools of this project in cache_invalidated_by. Add cacheable: true.

TOOL_CONFIRMATION_PHRASE_NOT_ACTION

Severity: warning Meaning. The tool requires confirmation and has no consent_action, so the confirmation question is built from the first sentence of its description (“Shall I <description> …?”). That sentence is a noun phrase or opens with a third-person verb, so the customer reads a broken question such as “Shall I synthetic transfer service for 25, Alex?”. Fix. Start the description with the action as a verb, for example “Transfer the amount to the beneficiary.”, or set consent_action on the confirmation.

TOOL_CONTEXT_ACCESS_READ_BEFORE_WRITE

Severity: warning Meaning. A tool’s CONTEXT_ACCESS.read names a key that only that same tool’s own ON_RESULT or ON_ERROR SET ever writes. That key cannot exist yet on the tool’s first call in a session, so the runtime’s context-injection guard fails the call closed every time (by design, an authorized-but-missing declared read never runs silently degraded). The declaration compiles clean today with no signal that the tool can never succeed on its first invocation. Fix. Either drop the key from CONTEXT_ACCESS.read if the tool does not actually need it as an input, or arrange for another tool or step earlier in the flow to populate it before this tool first runs. If the value truly can only come from this tool itself, read it from the tool result directly instead of declaring it as a CONTEXT_ACCESS dependency.

TOOL_OUTPUT_SCHEMA_INVALID

Severity: error Meaning. An MCP tool OUTPUT_SCHEMA is the JSON Schema of the tool result. A value that is not valid JSON cannot describe the result, so it is dropped. Fix. Write OUTPUT_SCHEMA as a single JSON object, for example {“type”: “object”, “properties”: {…}}.

TOOL_VOICE_UNKNOWN_KEY

Severity: warning Meaning. A tool-level VOICE block sets the delivery policy of the reply generated after the tool returns. Acoustic voice settings belong to the reply, template or channel. Fix. Keep only BARGE_IN and STREAMING in the tool VOICE block. The listening mode is fixed per call: set CONVERSATION.listening.mode on the agent instead.

UNKNOWN_SURVEY_TEMPLATE

Severity: error Meaning. A FEEDBACK block names a survey template that does not exist in the project being compiled. A contextual project build reports this as an error and cannot publish the unresolved reference. The template supplies the question and rating control; Feedback Insights selects a compatible scoring method when results are read. Fix. Correct the FEEDBACK_TEMPLATE name to one the project defines, the message lists every available template, or create the template in the project Surveys area. Names are compared exactly. If the survey was meant to be authored inline, remove FEEDBACK_TEMPLATE and give the block its own prompt and type instead.

UNSAFE_SIP_AUTHORITY_PATTERN

Severity: error Meaning. A call_transfer tool declares an approved_authorities entry whose VDN (userinfo) uses a pattern outside the constrained digit-class DSL, a general regex metacharacter, group, alternation, anchor, or an unbounded length, that compileApprovedAuthority rejects as unsafe rather than silently narrowing it. Fix. Express the VDN as an exact digit string or the digit-class DSL only (‘[0-9]’ single-digit ranges, for example, ‘sip:5210[0-9][0-9]@172.0.41.7:5060’). Regex syntax such as ’.*’, ’+’, ’|’, ’(…)’, or anchors is not supported.

UNTRUSTED_SIP_TRANSFER_DESTINATION

Severity: warning Meaning. A call_transfer tool declares a sipTransferId default without default_source: config or system, so the value is treated as model-supplied and ignored at transfer time. Fix. Add default_source: config and use a literal SIP URI when the agent should own the transfer destination. Otherwise remove the default and configure the target at the project level.

VOICE_BARGE_IN_ILLEGAL_PLACEMENT

Severity: error Meaning. Scoped barge-in changes a single playback lifecycle and is not a behavior-profile, channel-override, free-standing ON_START or YAML execution.voice setting. ABL EXECUTION.voice and MESSAGES never applied it, so they ignore it with the warning VOICE_DELIVERY_KEY_IGNORED_PLACEMENT instead. Fix. Move BARGE_IN to a template, FLOW step, GATHER field, tool VOICE block, or the VOICE block under RESPOND.

VOICE_DELIVERY_KEY_IGNORED_PLACEMENT

Severity: warning Meaning. BARGE_IN and STREAMING choose how one spoken output is delivered, so they take effect on a template, FLOW step, GATHER field, tool VOICE block or the VOICE block under RESPOND. Under ABL EXECUTION.voice and MESSAGES they have never had an effect: the line is ignored, is not stored, is never read as message text, and the agent keeps compiling. Fix. Remove the line, or move it to a template, FLOW step, GATHER field, tool VOICE block, or the VOICE block under RESPOND. For a call-wide default use CONVERSATION.listening.barge_in.

VOICE_INVALID_BARGE_IN

Severity: error Meaning. Per-output barge-in controls are a closed policy enum. An invalid value must not silently inherit or change playback behaviour. Fix. Set VOICE.BARGE_IN to allow or disallow, in lowercase.

VOICE_INVALID_STREAMING

Severity: warning Meaning. Per-output streaming is a boolean. A non-boolean value is ignored, so the next layer’s setting stands. Fix. Set VOICE.STREAMING to true or false.

VOICE_INVALID_TURN_DETECTION

Severity: error Meaning. turn_detection takes enabled (true or false) and an optional eou_threshold of 0.3, 0.4, 0.5, 0.6, 0.7, 0.8 or 0.9. An invalid value is dropped so the call keeps the next level’s setting. Fix. Write turn_detection with enabled: true or false, and eou_threshold set to 0.3, 0.4, 0.5, 0.6, 0.7, 0.8 or 0.9 (0.3 answers fastest, 0.9 Slow dictation waits longest).

VOICE_LISTENING_CALL_LEVEL_ONLY

Severity: error Meaning. The listening mode (background or turn_taking) is resolved once at call start and never changes during the call, so it cannot be authored per turn, step, template, field or tool. Fix. Remove LISTENING from this VOICE block and set CONVERSATION.listening.mode (agent or behavior profile) or the channel listeningMode instead.

VOICE_NO_INPUT_ILLEGAL_PLACEMENT

Severity: error Meaning. A caller-silence policy belongs to the step that waits for the caller. Under a GATHER field, RESPOND, branch, handler or response format it would never be applied, so it is rejected instead of being dropped. Fix. Move ON_NO_INPUT to the VOICE block directly under the FLOW step, or to CONVERSATION.listening.on_no_input for the whole agent.

VOICE_SIP_HEADERS_ILLEGAL_PLACEMENT

Severity: warning Meaning. SIP headers authored under VOICE are sent on the SIP BYE that ends the voice call, and only a COMPLETE condition ends the call that way. Under a RESPOND, FLOW step, GATHER field, branch, handler or response format there is no BYE to carry them, so the block is ignored. Fix. Move SIP_HEADERS to the VOICE block of the COMPLETE condition that ends the call. Headers for a transfer belong in the ESCALATE routing (routing.voice.sip_headers).

VOICE_STREAMING_LEGACY_PLACEMENT

Severity: warning Meaning. Per-output streaming belongs on a template, FLOW step, GATHER field, tool or concrete RESPOND output. Behavior-profile, channel-override and free-standing ON_START VOICE blocks, and YAML execution.voice, still accept it so existing agents keep compiling. ABL EXECUTION.voice and MESSAGES ignore it instead (see VOICE_DELIVERY_KEY_IGNORED_PLACEMENT). Fix. Move STREAMING to a template, FLOW step, GATHER field, tool VOICE block, or the VOICE block under RESPOND.

VOICE_TURN_DETECTION_WITH_END_OF_SPEECH

Severity: warning Meaning. While gateway turn detection is on, the gateway decides when the caller has finished, so the end-of-speech hint in the same block has no effect on that turn. Fix. Remove end_of_speech_turn_type / end_of_speech_timeout_ms from this block, or turn turn detection off here.