SamCDay/rune

1 Fork 0

Files

T

Nick f5ee557ee3 fix(CORRECTION-1/2): D2 producer contract + sentinel-tolerant reader; D3 telemetry wired into production path

ADAPT-1 baseline corrections (per handoff: fix D2+D3, record D1 as
FIXED_DEPTH_ROUTE_INAPPLICABLE — no OUTWARD semantics change).

CORRECTION-1 (D2): producer contract enforced
- Worker prompts updated: artifact file must contain exactly one JSON
  document; StageCompleted markers go to stdout ONLY
- _load_worker_artifact_file(): canonical reader requires pure JSON;
  LEGACY_COMPATIBILITY_ONLY permits exactly the known 'StageCompleted:
  <STAGE>' trailing sentinel (regex-anchored, no arbitrary salvage);
  corrupt JSON is never masked
- Acceptance: pure JSON OK; exact sentinel OK; arbitrary trailing text
  rejected; corrupt JSON rejected; all 6 stage prompts carry the contract

CORRECTION-2 (D3): StageAttempt telemetry wired into real production path
- research_event_relay: marks tool windows on tool running/completed/error/
  failed events (observational, never changes classification)
- StageExecutor._run_relay_with_telemetry: marks model windows around
  relay.run(); attempt threaded through _execute_stage into all 8 stage
  methods (telemetry_attempt kwarg, default None)
- PROVEN on a real production-path execution: ADVERSARIAL record shows
  wall=0.3006, model_time=0.3001 (1 window), tool_time=0.3001 (1 window) —
  overlapping by design, non-additive model respected
- telemetry failures never raise into the stage path (observational)

D1 NOT a code correction: recorded as
FIXED_DEPTH_ROUTE_INAPPLICABLE_TO_LOCAL_EVIDENCE_TASKS —
EXPECTED_ROUTING_LIMITATION / ADAPTIVE_REQUIREMENT. OUTWARD external-
evidence semantics preserved unchanged.

1163 tests pass (was 1158); ruff clean.

2026-09-09 22:18:50 +01:00

1.6 KiB

Raw Blame History

Research Stage: OUTWARD

Process capsule

Issue: {{ISSUE_NUMBER}}
Round: {{ROUND}}
Stage: OUTWARD (of 7)
Prior artifacts: {{PRIOR_ARTIFACTS}}
Blocking unknowns: {{BLOCKING_UNKNOWNS}}

Your task

Find and read enough reliable external evidence to answer the external research questions.

You MUST

DISCOVERY:

Issue real searches (webfetch of search engines, documentation sites)
Preserve query/tool receipts
Identify candidate sources with stable IDs (O1, O2, ...)
Classify each source (official docs, GitHub, academic, blog)
Select a bounded relevant set for reading

READING:

Actually fetch/read the selected sources
Extract evidence claims relevant to the subquestions
Record limitations and contradictions
Preserve source IDs and fetch/read receipts

You MUST NOT

Inspect unrelated Rune code
Design the Rune solution
Perform INNOVATION (that's the next stage)
Perform SYNTHESIS
Modify production code
Continue searching after the evidence requirement is satisfied
Treat search snippets as evidence (actually read the pages)

STOP CONDITION

Stop when each required external subquestion has:

Adequate evidence, OR
Explicit contradictory evidence, OR
A named unresolved gap

Output

Produce your stage artifact as structured text with:

DISCOVERY: queries issued, sources discovered with URLs
READING: sources actually read, evidence extracted
LIMITATIONS: what the sources don't cover
CONTRADICTIONS: where sources disagree

End with:

StageCompleted: OUTWARD  (stdout ONLY — the artifact file must contain exactly one JSON document, nothing else)

1.6 KiB Raw Blame History