Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/mattgierhart/PRD-driven-context-engineeringnpx agentmods add skills/mattgierhart/prd-driven-context-engineering/ghm-gate-checkWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/mattgierhart/prd-driven-context-engineering/ghm-gate-check)<a href="https://agentmods.dev/skills/mattgierhart/prd-driven-context-engineering/ghm-gate-check"><img src="https://agentmods.dev/badge/skills/mattgierhart/prd-driven-context-engineering/ghm-gate-check/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/mattgierhart/prd-driven-context-engineering/ghm-gate-check"><img src="https://agentmods.dev/badge/skills/mattgierhart/prd-driven-context-engineering/ghm-gate-check.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00067 | $0.01389 |
| Opus 5 | $0.00034 | $0.00694 |
| Sonnet 5 | $0.00013 | $0.00278 |
| Haiku 4.5 | $0.00007 | $0.00139 |
Grade A, and why
ghm-gate-check scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 150 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Gate Check
Validate whether the PRD stage is ready to advance to the next version. Delegates to the three-layer readiness scorer — SoT files → EPICs → stage — then surfaces the leverage view (what to fix first, and which EPICs it unblocks).
Workflow Overview
- Compute → run
scripts/readiness.py run --quietto refreshstatus/readiness.json - Read → parse
status/readiness.json - Report → PASS / WARN / BLOCK verdict with top blockers and causal links
- Recommend → actionable next steps (always highest-leverage first)
Authority
references/gate-criteria.md remains the canonical source of mandatory artifacts per gate. The scorer's GATE_REQUIREMENTS table mirrors it. Do not hand-roll checklists here — the scoring engine is the single source of truth.
Step 1: Compute
Run the orchestrator. It runs SoT → EPIC → stage in dependency order and writes status/readiness.json.
python scripts/readiness.py run --quiet
# exit 0 = all pass, 1 = warn, 2 = block, 3 = error
If the exit code is 3, report a runtime error and stop. If 0/1/2, proceed to Step 2.
Fallback: no scripts available
If scripts/readiness.py is missing or Python is unavailable, fall back to reading status/readiness.json directly. If that's also absent, report: "Readiness not yet computed — install scripts/requirements.txt and run python scripts/readiness.py run."
Step 2: Read
cat status/readiness.json
Extract:
summary.current_stage— the gate being evaluated and its scoresummary.top_blockers— ranked SoT files blocking progressstages.{target}— detailed stage block (dimensions, unmet_criteria, caps)epics.{id}— per-EPIC scores (cite the lowest ones)
Step 3: Report
Use this template. Fill every field from the JSON — do not improvise scores.
## Gate Check Report: {stage.gate_description}
**Verdict**: [PASS | WARN | BLOCK]
**Stage Score**: {stage.score} / 100 (warn < {threshold_warn}, block < {threshold_block})
**Date**: {now}
### Stage Dimensions
| Dimension | Score | Weight |
|-----------|-------|--------|
| required_ids_present | {score} | {weight} |
| relevant_sot_readiness | {score} | {weight} |
| cross_ref_integrity | {score} | {weight} |
| downstream_epic_readiness | {score or "n/a"} | {weight or "—"} |
### Top Blockers (leverage view)
1. **{file}** (score {score}) — blocks {N} EPICs: {EPIC-XX, …} — impact {impact}
2. …
### Unmet Criteria (high severity first)
- [high] {ref}: {reason}
- [medium] {ref}: {reason}
### Recommendation
**If PASS**: Advance to {next_version}. Run `ghm-status-sync` to update the README dashboard.
**If WARN / BLOCK**: Do not advance. Address top blockers in order — fixing the highest-impact SoT file cascades up the graph.
**Next action**: {top_blockers[0] → concrete fix}
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 150 lines · 67 tokens per session scan A beab4fc29e44
ghm-gate-check is a skill published in the GitHub repository mattgierhart/PRD-driven-context-engineering (182 stars, last pushed 8d ago), licensed MIT. It adds 67 tokens to every session and 1,389 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
ralplan
Consensus planning entrypoint that auto-gates vague ralph/autopilot/team requests before execution.
remember
Review reusable project knowledge and decide what belongs in project memory, notepad, or durable docs.
embedding-strategies
Select and optimize embedding models for semantic search and RAG applications. Use when choosing embedding models, implementing chunking strategies, or optimizing embedding quality for specific domains.
docx-comment-reply
Reply to comments (批注) in Word .docx/.doc files: extract comment context, draft replies, write threaded replies back, and validate OOXML.
explaining-machine-learning-models
Explain trained machine learning models through feature attribution, local explanations, and behavior summaries. Use as an explicit/manual helper once a model already exists, not for training ownership, leakage auditing, or general ML strategy selection.
aiwg-regenerate-copilot
Regenerate copilot-instructions.md for GitHub Copilot with vendor-specific content only.