Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add beevibe-ai/beevibe --skill beevibe-verify-prgit clone --depth 1 https://github.com/beevibe-ai/beevibeWrote 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/beevibe-ai/beevibe/beevibe-verify-pr)<a href="https://agentmods.dev/skills/beevibe-ai/beevibe/beevibe-verify-pr"><img src="https://agentmods.dev/badge/skills/beevibe-ai/beevibe/beevibe-verify-pr/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/beevibe-ai/beevibe/beevibe-verify-pr"><img src="https://agentmods.dev/badge/skills/beevibe-ai/beevibe/beevibe-verify-pr.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 | $0.00172 | $0.01468 |
| Opus 5.5 | $0.00069 | $0.00587 |
| Sonnet 5.5 | $0.00034 | $0.00294 |
| Haiku 4.5 | $0.00017 | $0.00147 |
Grade A, and why
beevibe-verify-pr 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 1mo 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 — 111 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Verify PR CI
When this fires
Any task session that ends with the agent having opened or pushed to a pull request. You are about to call mcp__beevibe__update_progress(done). STOP. Verify CI first.
The agent's natural exit ramp is "I pushed the PR, I'll mark done." That's wrong if CI is red or still pending — the PR can't merge, the task isn't actually delivered, and the human reviewer will bounce it right back. Run this skill so the green/red call is settled before you exit the session.
When this does NOT fire
- The deliverable isn't a PR (
document/analysis/report/design/artifact/previewwork_product types). CI doesn't apply. - You're in a chat session. The skill is for tracked task sessions only.
- The task is
failedor being reported viaupdate_progress(failed). CI doesn't gate failure reporting; just record the work and exit. - You're calling
report_blocker(notupdate_progress(done)). The PR isn't your blocker; the upstream dependency is.
The protocol
1. Identify the PR URL
You already have it — it's the work product you just recorded with create_work_product(type='pull_request', url='...'). Or pull it from the most recent push output if you haven't recorded the work_product yet (record it before verifying — the human reviewer needs it visible regardless of CI outcome).
If you can't find a PR URL, you didn't actually open a PR; skip this skill and reassess what you delivered.
2. Watch CI
timeout 600 gh pr checks "$PR_URL" --required --watch
EC=$?
--required filters to checks that are gating per the repo's branch protection rules; advisory/optional checks are skipped. --watch polls every 10 seconds until checks resolve. timeout 600 caps the wait at 10 minutes so you don't sit indefinitely if a runner stalls.
3. Branch on the exit code
$EC |
Meaning | Next step |
|---|---|---|
0 |
All required checks passed | Proceed to update_progress(done) |
1 |
One or more required checks failed | See 3a below |
124 |
timeout fired — checks still pending after 10m |
See 3b below |
| other | Unexpected (e.g. gh CLI error, no PR found, auth issue) |
Investigate; don't paper over with a retry loop |
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.
- 1mo ago First seen · 111 lines · 172 tokens per session scan A 1eade316a6b7
beevibe-verify-pr is a skill published in the GitHub repository beevibe-ai/beevibe (345 stars, last pushed yesterday), licensed Apache-2.0. It adds 172 tokens to every session and 1,468 once invoked, about $0.0007 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 2026-08-30.
Other skills, from other repositories
loop-engineering
A guide to building reliable loops for AI agents that plan, act, check their work, and continue until a task is truly complete. It includes independent verification and a proposer-reviewer pattern in which one agent's work is checked using outside evidence.
agent-evaluation
A guide to measuring whether an AI agent works reliably on the tasks that matter. It covers repeatable test environments, program checks, human-like scoring rules, failure analysis, and experiments that separate model problems from system problems.
bad-case-to-dpo
A guide to turning failed production agent runs into training and evaluation data. It explains how to find the first mistake, create preference examples for DPO (Direct Preference Optimization), and check that a fix does not damage other tasks.
eval-dataset-design
A guide to creating reliable test sets for AI agents. Each evaluation task includes the request, the simulated user's behaviour, a reset starting state, and a success check that can be independently verified.
kungfu-agent-onboarding
Discover the exact Kungfu Project, WorkConsole, WorkRef, Skill catalog, and Core Work state admitted to this Amp process.
kungfu-agent-onboarding
Use when a user asks to understand, start, inspect, extend, or safely operate installed Kungfu; verify the installed pack, select one intent route, personalize the explanation, and propose one smallest safe next action.