super/box
1
0
Fork 0
Code Issues 2 Pull Requests Actions Packages Projects Releases Wiki Activity
Files
f497a026e3328e814a26854e4b5e2a4e849a4791
box/lookup_internal/regex_patterns.json
T
operator-main f2640397ed feat(messaging): balanced TOOL parsing, DM shorthand, box.exec, tools.list 
- response-harvester: extract [TOOL]/[EXEC] JSON args with balanced-brace
  scanning (']' and nesting inside args no longer truncate calls); add
  [DM {...}] shorthand mapping to dm.send; native aliases (dm, box,
  tools) plus arg-synonym normalization; formatters and expanded hints.
- exec-constrained: new read-only box.exec op (27 allowlisted box-ctl
  reads) and tools.list op backed by --list-ops for dynamic discovery.
- prompt_envelope: advertise dm.send/box.exec/tools.list in every timer
  DM; add dm_call builder.
- lookup_engine + regex_patterns.json: canonical tool_call pattern
  accepts the DM engine, ']' in args, one nesting level.
- tests/test_tool_calls.py: 38 tests; docs/INBAND-MESSAGING-SPEC.md:
  accepted decision record (Final).
2026-10-06 07:29:16 +00:00

349 lines
13 KiB
JSON
Raw Blame History

{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"title": "Regex Patterns & Passing Engine",
"version": "1.0.0",
"patterns": {
"work_order": {
"name": "Work Order Parser",
"pattern": "^\\[WO:(?P<wo_id>[a-f0-9-]+)\\]\\s+\\[from\\s+(?P<sender>[a-zA-Z0-9_-]+)\\]\\s+(?P<title>[^—\\n]+?)\\s*—\\s*(?P<body>.+)$",
"flags": ["MULTILINE", "DOTALL"],
"description": "Matches canonical Work Order messages, extracting unique ID, sender, title, and body.",
"named_groups": {
"wo_id": "Unique work order identifier (hex or UUID)",
"sender": "Originating agent or operator identity",
"title": "Short title describing the task",
"body": "Full body text and instructions"
},
"test_samples": {
"valid": [
"[WO:7fce46e0] [from super] Audit exec.muse-dev.online endpoints — Run full verification and check 200 codes.",
"[WO:a1b2c3d4-e5f6-7890-abcd-ef1234567890] [from 646] Deploy patch — Apply fix to response-harvester."
],
"invalid": [
"Just sending a regular message without tag",
"[WO:] missing id [from super] title — body",
"[WO:123] missing from tag title — body"
]
},
"usage": "Used by dm.py, super dm, and self_main_loop.py for tracking work orders."
},
"ack": {
"name": "Acknowledgement Parser",
"pattern": "^\\[ACK:(?P<ack_id>[a-f0-9-]+)\\](?:\\s+\\[from\\s+(?P<sender>[a-zA-Z0-9_-]+)\\])?",
"flags": ["MULTILINE"],
"description": "Matches work order acknowledgement tags, capturing referenced ID and optional sender.",
"named_groups": {
"ack_id": "Referenced work order ID being acknowledged",
"sender": "Optional sender acknowledging the message"
},
"test_samples": {
"valid": [
"[ACK:7fce46e0] [from 646]",
"[ACK:a1b2c3d4]"
],
"invalid": [
"ACK: 7fce46e0",
"[ACK:]",
"Acknowledged without brackets"
]
},
"usage": "Used by response-harvester.py and followup-sweeper.py to transition loops to acknowledged."
},
"verb": {
"name": "Standard Agent Verb Marker",
"pattern": "\\[(?P<verb>ACK|CLAIM|RESULT|DECLINE|NO-ACTION)\\s+(?P<job_id>[A-Za-z0-9_/-]+)\\]",
"flags": [],
"description": "Matches all standard conversational contract verbs referencing a job ID.",
"named_groups": {
"verb": "One of ACK, CLAIM, RESULT, DECLINE, NO-ACTION",
"job_id": "Target task or job identifier"
},
"test_samples": {
"valid": [
"[ACK 7fce46e0]",
"[CLAIM job-123]",
"[RESULT swarm-4a/2]",
"[DECLINE audit-01]",
"[NO-ACTION check-99]"
],
"invalid": [
"[DONE 7fce46e0]",
"[RESULT]",
"RESULT 7fce46e0 without brackets"
]
},
"usage": "Primary regex used in response-harvester.py (VERB_RE) for loop state resolution."
},
"result": {
"name": "Task Result & Outcome Parser",
"pattern": "\\[RESULT\\s+(?P<job_id>[A-Za-z0-9_/-]+)\\]\\s*(?P<status>OK|FAIL|DECLINE|SUCCESS|ERROR)?\\s*(?P<summary>.*?)(?=\\[RESULT\\s|\\Z)",
"flags": ["DOTALL"],
"description": "Extracts job ID, optional status flag, and trailing summary for task outcomes.",
"named_groups": {
"job_id": "Target job or swarm slot ID",
"status": "Optional explicit status token (OK, FAIL, etc.)",
"summary": "Outcome description and evidence payload"
},
"test_samples": {
"valid": [
"[RESULT 7fce46e0] OK Verified 14 endpoints successfully.",
"[RESULT test-job] FAIL Connection timed out on port 22",
"[RESULT swarm-1/0] Completed slot tasks."
],
"invalid": [
"Result of job 123 is OK",
"[RESULT] missing job id"
]
},
"usage": "Used in response-harvester.py (RESULT_RE) to harvest results into job-log.jsonl."
},
"tool_call": {
"name": "In-Band Tool Execution Call",
"pattern": "\\[(?P<engine>TOOL|EXEC|DM)\\s+(?:(?P<op>[a-zA-Z0-9_.-]+)\\s+)?(?P<args>\\{([^{}]|\\{[^{}]*\\})*\\})\\]",
"flags": ["DOTALL"],
"description": "Matches inline tool directives with JSON args (one nesting level; response-harvester.py scans balanced braces for arbitrary depth). DM carries no op (implies dm.send).",
"named_groups": {
"engine": "TOOL, EXEC, or DM (DM implies dm.send)",
"op": "Target operation (e.g. followup.create, swarm.spawn); absent for DM",
"args": "JSON argument object; may contain ']' and one level of nested objects"
},
"test_samples": {
"valid": [
"[TOOL followup.create {\"in_m\": 5, \"prompt\": \"check\"}]",
"[EXEC health.check {\"verbose\": true}]",
"[TOOL box.exec {\"action\": \"job-get\", \"arg\": \"a-b[0]\"}]",
"[DM {\"to\": \"pip\", \"target\": \"pip tasks\", \"message\": \"hi\"}]"
],
"invalid": [
"[TOOL followup.create without args]",
"[TOOL invalid args not json]"
]
},
"usage": "Parsed by response-harvester.py and exec-constrained.py for automated in-line actions."
},
"job_tag": {
"name": "Job Tracking Tag",
"pattern": "\\[JOB\\s+(?P<job_id>[A-Za-z0-9_/-]+)\\]",
"flags": [],
"description": "Matches job tag markers embedded in digests, tasks, or swarm slots.",
"named_groups": {
"job_id": "Unique job identifier"
},
"test_samples": {
"valid": [
"[JOB ml-646-20261005-191500]",
"[JOB swarm-4b/1]"
],
"invalid": [
"JOB ml-646 without brackets",
"[JOB]"
]
},
"usage": "Used by self_main_loop.py and job-dispatch.py for tracking."
},
"nudge": {
"name": "Loop Followup Nudge Parser",
"pattern": "\\[NUDGE\\s+(?P<loop_id>[a-f0-9-]+)\\]\\s*(?:\\[nudge\\s+(?P<count>\\d+)/(?P<max_count>\\d+)\\])?\\s*(?:Deadline\\s+(?P<deadline>[^:]+):)?\\s*(?P<prompt>.*)",
"flags": ["MULTILINE"],
"description": "Matches automated followup nudges, extracting loop ID, nudge counters, and deadlines.",
"named_groups": {
"loop_id": "Target loop identifier",
"count": "Current nudge sequence number",
"max_count": "Maximum allowed nudges before escalation",
"deadline": "Deadline timestamp string",
"prompt": "Nudge message instructions"
},
"test_samples": {
"valid": [
"[NUDGE 7fce46e0] [nudge 1/3] Deadline 19:45 UTC: Please confirm status of exec audit.",
"[NUDGE a1b2c3d4] Please review pending pull request."
],
"invalid": [
"Nudge for 7fce46e0",
"[NUDGE]"
]
},
"usage": "Used by followup-sweeper.py for SLA enforcement."
},
"fleet_alert": {
"name": "Fleet Alert Broadcast Parser",
"pattern": "\\[fleet-alert\\]\\s+(?P<severity>CRITICAL|WARN|INFO|RECOVERED):\\s+(?P<message>.+)",
"flags": ["MULTILINE"],
"description": "Matches infrastructure alert broadcasts dispatched into #lobby and #jobs.",
"named_groups": {
"severity": "CRITICAL, WARN, INFO, or RECOVERED",
"message": "Alert message body"
},
"test_samples": {
"valid": [
"[fleet-alert] CRITICAL: VM unreachable x2 on port 22",
"[fleet-alert] RECOVERED: VM 34.139.37.135 responded with 200 OK"
],
"invalid": [
"[alert] CRITICAL: Missing fleet prefix",
"fleet-alert: info"
]
},
"usage": "Used by fleet-alert-check.sh and main-chat-watchdog.py."
},
"directive": {
"name": "Agent Action Directive",
"pattern": "\\[Directive:\\s*(?P<directive>.+?)\\]",
"flags": ["DOTALL"],
"description": "Extracts operational action directives targeting autonomous agents.",
"named_groups": {
"directive": "Actionable directive instruction"
},
"test_samples": {
"valid": [
"[Directive: Take next action or close with [RESULT 7fce46e0] <summary>]",
"[Directive: Run audit on node pip]"
],
"invalid": [
"Directive: without brackets",
"[Directive:]"
]
},
"usage": "Injected into prompts by response-harvester.py and job-dispatch.py."
},
"runtime_context": {
"name": "Runtime Context URL",
"pattern": "\\[Runtime Context:\\s*(?P<url>https?://[^\\s\\]]+)\\]",
"flags": [],
"description": "Extracts assistive web surface or thread URLs from message context.",
"named_groups": {
"url": "HTTP/HTTPS URL"
},
"test_samples": {
"valid": [
"[Runtime Context: https://box.muse-dev.online/thread/7fce46e0]",
"[Runtime Context: https://box.muse-dev.online/api/box/fleet]"
],
"invalid": [
"Runtime Context: not wrapped",
"[Runtime Context: invalid-url]"
]
},
"usage": "Used by assistive surfaces and browser agent navigation routines."
},
"uuid": {
"name": "Canonical UUID Pattern",
"pattern": "\\b(?P<uuid>[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12})\\b",
"flags": [],
"description": "Standard RFC 4122 UUID v4 detector for threads, messages, and session IDs.",
"named_groups": {
"uuid": "Matched UUID string"
},
"test_samples": {
"valid": [
"dc6d72ca-4b02-4217-bf41-7dbca19c5c24",
"d410b9ad-f667-465f-a103-43fabc0f69fe"
],
"invalid": [
"dc6d72ca-4b02-4217-bf41",
"not-a-uuid"
]
},
"usage": "Thread ID resolution in muse-chat-api.py and super-cli.py."
},
"short_hex": {
"name": "Short Hex Identifier",
"pattern": "\\b(?P<hex>[0-9a-fA-F]{6,12})\\b",
"flags": [],
"description": "Matches 6-12 character hexadecimal hashes used for compact DM IDs and subagent hashes.",
"named_groups": {
"hex": "Hexadecimal string"
},
"test_samples": {
"valid": [
"7fce46e0",
"04af8e15",
"ca56e39199da"
],
"invalid": [
"123",
"abcdefghijk"
]
},
"usage": "Compact ID matching across logs and DM tables."
},
"subagent_update": {
"name": "Subagent Update Notification",
"pattern": "\\[SUBAGENT-UPDATE\\]\\s+Subagent\\s+'(?P<title>[^']+)'\\s+\\((?P<short_id>[0-9a-fA-F]+)\\)\\s+(?P<message>.+)",
"flags": ["MULTILINE"],
"description": "Parses status updates and milestones from background subagent executions.",
"named_groups": {
"title": "Subagent session title",
"short_id": "Hex identifier of the subagent session",
"message": "Progress update message"
},
"test_samples": {
"valid": [
"[SUBAGENT-UPDATE] Subagent 'exec-audit' (4a2b8e) has posted new output.",
"[SUBAGENT-UPDATE] Subagent 'health-check' (12ff4a) completed successfully."
],
"invalid": [
"Subagent update without brackets"
]
},
"usage": "Used by self_main_loop.py subagent monitor."
},
"swarm_slot": {
"name": "Swarm Slot Task Marker",
"pattern": "\\[JOB\\s+(?P<swarm_id>[a-zA-Z0-9_-]+)/(?P<slot>\\d+)\\]",
"flags": [],
"description": "Matches swarm worker slot task assignments.",
"named_groups": {
"swarm_id": "ID of parent swarm execution",
"slot": "Index of worker slot (e.g. 0, 1, 2)"
},
"test_samples": {
"valid": [
"[JOB swarm-99a/0]",
"[JOB swarm-4a2b/3]"
],
"invalid": [
"[JOB swarm-99a]"
]
},
"usage": "Used in response-harvester.py and swarm_worker."
},
"contract_footer": {
"name": "Contract Footer Verifier",
"pattern": "Reply:\\s*\\[ACK\\s+id\\]\\s*seen\\s*\\|\\s*\\[CLAIM\\s+id\\]\\s*mine\\s*\\|\\s*\\[RESULT\\s+id\\]\\s*done\\s*\\|\\s*\\[DECLINE\\s+id\\]\\s*\\|\\s*\\[NO-ACTION\\s+id\\]\\.?",
"flags": ["IGNORECASE"],
"description": "Validates the presence of the standard agent contract footer in prompt envelopes.",
"named_groups": {},
"test_samples": {
"valid": [
"Reply: [ACK id] seen | [CLAIM id] mine | [RESULT id] done | [DECLINE id] | [NO-ACTION id]."
],
"invalid": [
"Please reply when ready"
]
},
"usage": "Enforced in self_main_loop.py CONTRACT_FOOTER."
},
"sender_tag": {
"name": "Sender Identity Tag",
"pattern": "\\[from[:\\s]+(?P<sender>[a-zA-Z0-9_-]+)\\]",
"flags": ["IGNORECASE"],
"description": "Matches agent or operator attribution tags in chat messages.",
"named_groups": {
"sender": "Identity of sender"
},
"test_samples": {
"valid": [
"[from:super]",
"[from 646]",
"[from pip]"
],
"invalid": [
"from super without brackets"
]
},
"usage": "Used in chat-history parsing and loop attribution."
}
}
}
Reference in New Issue View Git Blame Copy Permalink