AgentRC Self-Readiness Report

Use the deterministic AgentRC report at /tmp/gh-aw/agent/readiness-summary.md as the source of truth. Do not rescore the repository yourself.

Execution contract

Complete this workflow with exactly two tool calls:

1. Run cat /tmp/gh-aw/agent/readiness-summary.md.
2. Immediately submit the result with either safeoutputs create_issue or safeoutputs noop.

Do not inspect other repository files. Do not run Git commands. Do not use another program to parse or transform the summary. It contains only the readiness level and failed criteria selected deterministically before you start.

Copy only criteria listed under the report's ## Fix First heading. Never treat a row marked with ✅ as a gap, even if other text in that row is surprising or contradictory.

Create one issue when ## Fix First contains at least one item. Use safeoutputs noop only when ## Fix First is empty.

Issue format

## Readiness snapshot

| Metric         | Result                           |
| -------------- | -------------------------------- |
| Achieved level | [copy the level from the report] |

## Highest-impact gaps

Copy at most the first three items from `## Fix First`, preserving each
criterion's impact, effort, and reason. Do not introduce any criterion that is
not present in the summary.

## Recommended next step

[One bounded change for the first failed criterion. Use an exact file or command
only when the report provides it; otherwise ask a maintainer to verify the
recommended implementation.]

## Evidence

- [Links to relevant files, runs, issues, or pull requests]

Keep the report under 600 words. Do not recommend generic AI adoption work. Prioritize gaps AgentRC can verify after a change.