git:20260714.7c51b62 to git:20260921.41cb25e

1 added, 1 removed. Audit A to A.

---

name: fable-fact-check

description: Claim-by-claim verification of docs, status reports, launch claims, changelogs, READMEs, runbooks, and "done/tested/working/live" assertions against source, tests, artifacts, and runtime evidence. Use when the user asks Codex to fact-check, verify claims, audit truthfulness, or compare docs to reality.

---

# Fable Fact Check

Verify claims against actual evidence. Do not edit claims unless the user asks for a correction pass.

For ECF-style governed runs, use `../../references/ecf-run-contract.md`. For large or high-risk fact checks, use real Codex subagents when the runtime exposes a subagent tool and the user has not opted out; treat multi-document, launch/status, security/privacy/money/data/API, release-readiness, or many-claim checks as large by default. Otherwise run `single-agent multi-lens` and say why no subagents were used.

- For large or high-risk Fable tasks, recommend running the parent task on `gpt-5.6-sol` with Ultra (`model_reasoning_effort = "ultra"`) when available. Ultra may delegate proactively, but still explicitly request parallel delegation for disjoint Fable lenses when the runtime supports subagents; otherwise use `single-agent multi-lens` and report the reason.

+ For large or high-risk Fable tasks, use `gpt-5.6-sol` with Ultra (`model_reasoning_effort = "ultra"`) for the parent coordinator when it is selected or can be configured before the run. Record the requested and actual coordinator model and effort plus any fallback; never imply Sol Ultra was active when it was not. Request parallel delegation for disjoint Fable lenses. Use `gpt-5.6-luna` with `medium` reasoning for every delegated subagent by default, setting the model and effort explicitly on each spawn when the runtime supports it. Keep load-bearing synthesis and final verification with the Sol Ultra coordinator. If the runtime cannot honor the worker selection, report the requested and actual worker model plus the fallback; never claim Luna was used without runtime evidence. Otherwise use `single-agent multi-lens` and report the reason.

## Workflow

1. Identify the document, status report, or claim set.

2. Restate authority boundaries and declare the ECF run mode: `multi-agent` only with real spawned subagents, otherwise `single-agent multi-lens`.

3. Extract checkable claims. Prioritize words like done, shipped, live, tested, working, complete, supports, secure, verified, deployed, public, and automated.

4. For each claim, find the implementing source, tests, artifacts, deployment config, or runtime surface that would make it true.

5. Classify each claim:

- true

- false

- partial

- stale

- unverifiable

6. Cite exact evidence for every classification.

7. Separate docs/source mismatch from environment/tooling failure.

8. If asked to fix docs, patch only the claims that evidence disproves or bounds.

## Evidence Safety

Never print raw secrets, tokens, private keys, wallet keys, credential files, or `.env` values. Redact secret-like values and cite only the file/path/key name needed to explain the issue.

## Output

Use a compact table when many claims are checked:

```text

Claim | Verdict | Evidence | Correction

```

End with the highest-risk false or partial claims and the safest next verification step.

Include a compact `Workflow Trace` with mode, ECF contract status, lenses covered, spawned agents or no-subagent reason, verification method, and coverage gaps.