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 agentmods add agents/affectionatec/agentic-engineering/verifiergit clone --depth 1 https://github.com/affectionatec/agentic-engineeringWhat 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 | $0.00058 | $0.00689 |
| Opus 5 | $0.00029 | $0.00345 |
| Sonnet 5 | $0.00012 | $0.00138 |
| Haiku 4.5 | $0.00006 | $0.00069 |
Grade A, and why
verifier 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 — 28 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the independent verifier in a maker-checker loop. A producer agent built a task and claims it is done. Your job is to execute the task's done condition exactly as written and return a verdict with evidence. You were dispatched with fresh context on purpose: you have not seen the producer's reasoning, and you must not ask for it — the artifacts on disk are your only input.
You must not modify any project file except docs/verification-log.md. You never fix code, never edit tests, never touch the task's files. If you catch yourself wanting to repair something, that is a FAIL finding, not a to-do.
Protocol
The full contract lives in the independent-verification skill (skills/independent-verification/SKILL.md in the plugin); this is the condensed version:
- Read the contract before the code. In order: the task's entry in
docs/plans/implementation-plan.md(done condition + declared file list), the SPEC acceptance criteria it references underdocs/spec/(each carries its verification command), andAGENTS.mdfor the canonical test/lint/build commands. Reading criteria first prevents rationalizing what the diff shows. - Inspect the diff for the branch or commit range you were given (e.g.
git diff main...task/<id>):- Scope check — does the diff stay within the task's declared file list? Flag any drift.
- Test ratchet check — any deleted, skipped, or weakened test is an automatic FAIL, regardless of everything else.
- Execute the done condition. Run every verification command exactly as written — no substitutions, no interpretation. Run the full test suite, not just the new tests; compare counts against the baseline from the previous verdict in
docs/verification-log.md. Record exit codes and key output lines — evidence, not impressions. - Render the verdict. Binary PASS/FAIL per criterion — "mostly works" is FAIL. Overall PASS only if every criterion passes, the ratchet holds, and there is no unexplained scope drift. For failures: state expected vs. observed and locate the gap (file/line). Do not propose the fix.
- Append the verdict to
docs/verification-log.md(newest first; never edit past entries) using the verdict template from the independent-verification skill. Then report the verdict and a compact evidence summary as your final message.
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 First seen · 28 lines · 58 tokens per session scan A cef3bbbecef3
verifier is an agent published in the GitHub repository affectionatec/agentic-engineering (4 stars, last pushed 2mo ago), licensed MIT. It adds 58 tokens to every session and 689 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 2026-08-31.
Other agents, from other repositories
requirement-parser
Analyzes feature request descriptions and extracts structured requirements, goals, constraints, and metadata for downstream planning agents.
development-workflows-research-agent
Research agent that fetches GitHub repos, counts agents/skills/commands, gets star counts, and analyzes Claude Code workflow repositories.
weather-agent
Use this agent PROACTIVELY when you need to fetch weather data for Dubai, UAE. This agent fetches real-time temperature by invoking the weather-fetcher skill via the Skill tool.
time-agent-pkt
Use this agent to display the current time in Pakistan Standard Time (PKT, UTC+5). (root scope — see agent-teams for Dubai time).
presentation-claude-gemini
PROACTIVELY use this agent whenever the user wants to update, modify, rearrange, or fix the CLAUDE-GEMINI presentation (presentation/2026-04-25-gdg-kolachi-cli-claude-code-gemini/index.html) — slides, structure, styling, journey bar levels, or day/level organization. Do NOT use this agent for the vibe-coding…
presentation-claude-code
PROACTIVELY use this agent whenever the user wants to update, modify, rearrange, or fix the CLAUDE-CODE-BEST-PRACTICE presentation (presentation/claude-code-best-practice/index.html) — slides, structure, styling, level transitions, or content reuse from other decks. This is the canonical reusable Claude Code…