Index of every page on this site: llms.txt
All skills
google avatar

/egm-review-loop

@70406c4
by googlegoogle/elephant-goldfish-model22k stars
1,589
Run it yourself

Executes the Antagonistic-Review process from the Elephant-Goldfish Model (EGM) via an automated multi-agent review loop to audit documents against qualitative standards and structural templates, and automatically apply approved fixes. Use when reviewing design docs, READMEs, feature specifications, implementation plans, or verifying document completeness and tone.

Use this Skill: https://skilld.dev/gh/google/elephant-goldfish-model/egm-review-loop

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

TEST.md

≈295 tokens on demand. Your agent reads this file only when SKILL.md points to it.

Test Plan: Goldfish Review Loop Skill

Frontmatter and link checks run in pytest tests/unit (see TESTING.md).

Protocol Verification Checklist

When evaluating or exploring egm-review-loop interactively in chat, verify:

  • Step 1 (Intake & Config): Validates required inputs, resolves standard paths, and saves each URL in GoldfishContext as a file before the first round, asking for a local copy of any page that cannot be saved.
  • Step 2.1 (Target Resolution): Unconditionally re-scans line numbers before auditing.
  • Step 2.2 (Review Swarm): Spawns subagents (or sequentially iterates personas in degraded mode) across requested standards.
  • Step 2.3 & 2.4 (Resolve Handoff & Turn Boundary): Hands findings to egm-resolve and respects human decision gates.
  • Step 2.4 (Round Control): Keeps contested and accepted findings settled, re-opens fixed findings that reviewers raise again, judges clean rounds before applying fixes, records failed reviewers (and reviewers with no status in the synthesizer summary) as not run, stops after two consecutive clean rounds, and asks the user at the 5-round cap or on churn.

Source: SKILL.md on GitHub

No rule matched.

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

1 warning15d3 checks · Risk SAFE
  • Gen Agent Trust Hub16d

    This skill facilitates a structured, multi-agent document review and repair loop. It includes security considerations such as the ingestion of external data and the retrieval of content from user-provided URLs. While these patterns warrant review, they are used within the skill's intended functionality to provide context for document auditing. See detailed analysis for context.

  • Socket15d

    No alerts

  • Snyk16d

    Risk: MEDIUM · 1 issue

Signed by skilld at 70406c4. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 3 days ago.

Activeupdated 2 weeks ago
disable-model-invocation
true

README badge

README badge for google/elephant-goldfish-model/egm-review-loop