GitHub Agentic Workflows is a GitHub CLI extension that lets developers define AI-assisted repository automation in Markdown and run it through GitHub Actions. It is intended for tasks requiring interpretation or reasoning, such as issue triage, pull-request review, CI investigation, documentation maintenance, and dependency analysis. The catalogue entries provide skills and agents for working with these workflows.
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/github/gh-awnpx agentmods add skills/github/gh-aw/model-routing-goldenWrote 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/github/gh-aw/model-routing-golden)<a href="https://agentmods.dev/skills/github/gh-aw/model-routing-golden"><img src="https://agentmods.dev/badge/skills/github/gh-aw/model-routing-golden/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/github/gh-aw/model-routing-golden"><img src="https://agentmods.dev/badge/skills/github/gh-aw/model-routing-golden.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00015 | $0.00378 |
| Opus 5.5 | $0.00006 | $0.00151 |
| Sonnet 5.5 | $0.00003 | $0.00076 |
| Haiku 4.5 | $0.00002 | $0.00038 |
Grade A, and why
model-routing-golden 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 27d 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.
What it actually says
Model-Routing Golden Fixtures
Use this skill when a model-routing golden test fails, analysis output changes intentionally, or a new routed workflow run needs coverage.
Verify
Run make verify-model-routing-golden (or
python3 scripts/model-routing-golden.py --verify) before changing fixtures.
Treat a mismatch as a behavior change to investigate, not a reason to regenerate
immediately.
Refresh expected outputs
When analysis behavior intentionally changes, run
make update-model-routing-golden. Review the complete diff for each
expected*.json file and confirm that routing status, classifier attribution,
AWF credits, token totals, warnings, and subagent costs match the intended
behavior. Keep separate legacy goldens when missing artifacts cause documented
information loss. Do not regenerate outputs merely to silence an unexplained
failure.
Add or replace a fixture
Capture a run with:
python3 scripts/model-routing-golden.py --run <id-or-url> --case <case-name>
Use --replace only when intentionally updating an existing source run. The
tool downloads through the local gh aw logs artifact path, minimizes and
redacts its files, and refuses to write if analysis on the fixture differs from
analysis on the full download. Use --source-dir to exercise capture against
the synthetic downloaded-run fixture without GitHub access.
Add the case and its source scenario to modelRoutingGoldenCases and the
fixture README. Then run make update-model-routing-golden, inspect all
generated files, and run make verify-model-routing-golden. Retain the
established file layout and logs/ ignore exception so raw proxy-log fixtures
remain tracked.
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.
- 27d ago First seen · 47 lines · 15 tokens per session scan A 4153e11d1cd7
model-routing-golden is a skill published in the GitHub repository github/gh-aw (5,579 stars, last pushed yesterday), licensed MIT. It adds 15 tokens to every session and 378 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-10-11.
model-routing-golden compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| model-routing-goldengithub/gh-aw | Skill | 5.6k | yesterday | MIT | A | 15 | Claude Code, Codex |
| watch-prjustincampbell/gh-watch | Skill | 4 | 7d ago | MIT | A | 48 | Claude Code |
| watch-tagjustincampbell/gh-watch | Skill | 4 | 7d ago | MIT | A | 62 | Claude Code |
| watch-branchjustincampbell/gh-watch | Skill | 4 | 7d ago | MIT | A | 45 | Claude Code |
| watch-commitjustincampbell/gh-watch | Skill | 4 | 7d ago | MIT | A | 48 | Claude Code |
| c-githubdaxaur/openpaw | Skill | 177 | 5mo ago | MIT | A | 48 | Claude Code, Codex |
| update-architecture-docsAlexSkrypnyk/scaffold | Skill | 11 | yesterday | GPL-3.0 | A | 46 | Claude Code |