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 tamnguyendinh/Anvien --skill supervisorgit clone --depth 1 https://github.com/tamnguyendinh/AnvienWrote 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/tamnguyendinh/anvien/supervisor)<a href="https://agentmods.dev/skills/tamnguyendinh/anvien/supervisor"><img src="https://agentmods.dev/badge/skills/tamnguyendinh/anvien/supervisor/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/tamnguyendinh/anvien/supervisor"><img src="https://agentmods.dev/badge/skills/tamnguyendinh/anvien/supervisor.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 | $0.00051 | $0.01006 |
| Opus 5 | $0.00026 | $0.00503 |
| Sonnet 5 | $0.00010 | $0.00201 |
| Haiku 4.5 | $0.00005 | $0.00101 |
Grade A, and why
supervisor 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 — 90 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Supervisor Review
Use this skill to independently decide whether a claim, artifact, work result, or completion statement can be accepted in a repo/project.
The review target is only the entry point. Verify the real claim, write the report, then summarize the verdict.
Core Law
Zero-trust review: treat every claim, artifact, result, and completion statement as untrusted until independently verified against repo/project reality with sufficient evidence.
Role Boundary
Supervisor Review gives an acceptance verdict. It does not repair the work while reviewing unless the user explicitly asks for a separate implementation task.
Do not blend review and fix work. If review finds a problem, write a REJECT report with evidence and the required next step.
Approval Standard
PASS only when all are true:
- the real claim is clear;
- authority is identified and not blocking;
- source/project reality has been inspected where required;
- evidence proves the full claim, not a narrower claim;
- the affected invariant is closed for the reviewed scope;
- no required follow-up remains before acceptance.
REJECT when any are true:
- the claim is false, incomplete, unsafe, or misleading;
- evidence is missing, stale, indirect, partial, or narrower than the claim;
- source/project reality contradicts the claim;
- authority conflicts or is missing for acceptance;
- the fix only addresses the visible symptom while same-invariant surfaces remain unchecked or broken;
- any required action remains before acceptance.
Always Do
- State the real claim and authority before judging.
- Verify the full claim against repo/project reality before PASS.
- Inspect source before trusting build/test/report output when code reality matters.
- Use Anvien when codebase topology, impact, contracts, dependencies, or affected flows matter.
- Review the affected invariant, not only the visible symptom or changed lines.
- Include direct evidence, preferably file/line evidence when source is involved.
- Give exactly one verdict: PASS or REJECT.
- Apply feedback discipline: read, understand, verify, evaluate, then judge.
- Ask or REJECT when unclear scope prevents a sound verdict.
- State claims only with evidence gathered for the current review state.
What ships with it
4 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 · 90 lines · 51 tokens per session scan A 35bae1631cb6
supervisor is a skill published in the GitHub repository tamnguyendinh/Anvien (9 stars, last pushed yesterday), licensed MIT. It adds 51 tokens to every session and 1,006 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 2026-09-03.
Other skills, from other repositories
blast-radius-review
Review a change by its blast radius, using Kin's graph to find what downstream code the change can reach. Use when reviewing a diff or pull request, deciding whether an edit is safe, or answering "what else breaks if I change this" in a repository admitted to Kin.
qartez
Semantic code intelligence skill for qartez MCP. Use qartez tools for code exploration, impact analysis, and guarded refactors; treat built-in code-tool denials as routing signals and recover via qartez replacements.
security-review
Perform a focused security review of pending git changes to identify high-confidence security vulnerabilities with real exploitation potential. Use this skill when the user asks for a security review, security audit, vulnerability scan, or wants to check pending changes on a branch for security issues before merging.…
huggingface-llm-trainer
Train or fine-tune language models with TRL or Unsloth on Hugging Face Jobs, including SFT, DPO, GRPO, reward models, and GGUF conversion. Use for cloud LLM training; use huggingface-vision-trainer for vision tasks.
feature-dev
Guide a feature implementation through a structured seven-phase workflow with deep codebase understanding, clarifying questions, parallel architecture design, and quality review. Use this skill when the user asks to build a new feature, add functionality, or wants a methodical approach to implementation rather than…
review
Validate plans, execution, or PRs against wish criteria — returns SHIP / FIX-FIRST / BLOCKED with severity-tagged gaps.