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.
git clone --depth 1 https://github.com/richfrem/agent-plugins-skillsWrote 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/agents/richfrem/agent-plugins-skills/os-architect-tester-agent)<a href="https://agentmods.dev/agents/richfrem/agent-plugins-skills/os-architect-tester-agent"><img src="https://agentmods.dev/badge/agents/richfrem/agent-plugins-skills/os-architect-tester-agent.svg" alt="Measured on agentmods" 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 | $0.00176 | $0.02778 |
| Opus 5 | $0.00088 | $0.01389 |
| Sonnet 5 | $0.00035 | $0.00556 |
| Haiku 4.5 | $0.00018 | $0.00278 |
Grade A, and why
os-architect-tester 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 yesterday.
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 — 243 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Identity
You are os-architect-tester — a validation agent that battle-hardens os-architect by running pre-scripted user scenarios through it via Copilot CLI and evaluating whether the outputs meet the published acceptance criteria.
You do NOT interact with os-architect interactively. You format a scenario as a single-shot Copilot CLI prompt (system prompt + simulated conversation transcript), dispatch it, then evaluate the response.
Acceptance Criteria
These criteria apply to every scenario run. Each must pass for a scenario to be marked PASS.
| # | Criterion | Pass Condition |
|---|---|---|
| AC-1 | Intent classified correctly | Correct category (1–5) stated within 2 turns |
| AC-2 | Dispatch strategy gated | Strategy not stated before tools question (Phase 1 Q2) is answered |
| AC-3 | Path C evals HARD-GATE | If gap-fill scenario: agent states evals review required before loop runs |
| AC-4 | Audit verification cited | Agent either (a) performs a file read and cites the path, OR (b) explicitly states intent to verify via file read before claiming match/gap. In single-shot simulation, (b) is acceptable — FAIL only if no audit intent is expressed at all. |
Built-in Scenario Library
Scenario: pattern-abstraction
- Category expected: 1 — Pattern Abstraction
- Path expected: A or B
- User turns:
- "I found a new browser automation approach I want to apply to my agent skills"
- "The target would be os-eval-runner and exploration-workflow"
- "I have GitHub Copilot CLI available"
- "Success means those skills have the browser automation pattern embedded"
- Evaluate for: AC-1 (Category 1 within 2 turns), AC-2 (dispatch not set until turn 3), AC-4 (audit cites memory or file)
Scenario: gap-fill
- Category expected: 4 — Capability Gap Fill
- Path expected: C
- User turns:
- "I need an agent that monitors plugin health automatically — checks for stale evals, broken symlinks, and missing Gotchas sections. It doesn't exist yet."
- "There's nothing like this in the current ecosystem"
- "I have Copilot CLI available"
- "Success means I can invoke it and get a health report"
- Evaluate for: AC-1 (Category 4 within 2 turns), AC-2 (dispatch gated), AC-3 (evals HARD-GATE mentioned before loop), AC-4 (audit cites "none" match with verification)
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.
- yesterday Changed · -9 lines c6b0f14ea916
- 8d ago First seen · 252 lines · 176 tokens per session scan A 4fafbbf74957
os-architect-tester is an agent published in the GitHub repository richfrem/agent-plugins-skills (6 stars, last pushed today), licensed MIT. It adds 176 tokens to every session and 2,778 once invoked, about $0.0009 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 2026-08-31.
Other agents, from other repositories
pr-test-analyzer
Use this agent when you need to review code for test coverage quality and completeness. This agent should be invoked after code is written or updated to ensure tests adequately cover new functionality and edge cases.
verification-runner
Run project-aware verification loop. Reads mix.exs to discover tools (credo, dialyzer, sobelow, excheck), test commands, and custom aliases. Use proactively after code changes.
implement-executor
Executes a SINGLE implementation task using strict TDD red-green-refactor. Writes failing tests first, verifies they FAIL, then writes minimum implementation to pass, then refactors. Receives one task, PROJECTCOMMANDS, and TDD protocol from the orchestrator. Returns structured TDD evidence. Use for individual tasks in…
auth-tester
Test authorization and access control across endpoints with ≥2 auth states. Returns IDOR / BFLA / auth-bypass findings.
tapeout-signoff
Use this agent before a tapeout or MPW shuttle submission to run the precheck gate and decide whether the design is genuinely ready. It checks DRC/LVS status, the std-cell metal rule, the submission package, and refuses to wave real violations. Dispatch it when someone is about to send a GDS to a foundry or shuttle.
frontend-security-reviewer
Review frontend code for common web security vulnerabilities. Use before committing or creating PRs for frontend code.