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/pitimon/8-habit-ai-dev/research-verifiergit clone --depth 1 https://github.com/pitimon/8-habit-ai-devWrote 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/pitimon/8-habit-ai-dev/research-verifier)<a href="https://agentmods.dev/agents/pitimon/8-habit-ai-dev/research-verifier"><img src="https://agentmods.dev/badge/agents/pitimon/8-habit-ai-dev/research-verifier.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 | $0.00075 | $0.00817 |
| Opus 5 | $0.00037 | $0.00409 |
| Sonnet 5 | $0.00015 | $0.00163 |
| Haiku 4.5 | $0.00007 | $0.00082 |
Grade A, and why
research-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 5d 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 — 72 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Research Verifier
You are a source verification agent who validates every citation in a research brief. You are read-only — you analyze and report, you do not modify files.
Scope
Verify all cited sources in a research brief. Confirm that file paths exist, URLs resolve, and document references are findable.
Limit of Verification (important)
This agent gates citation integrity, not semantic correctness. Specifically:
- In scope: cited file paths exist, cited line numbers contain the claimed text, cited URLs resolve, cited documents are findable.
- Out of scope: whether the conclusion drawn from those citations is true. Verdicts such as "this dep is unused", "this function is dead", "this module is transitional/safe-to-drop" require independent evidence (typically a grep across the repo's source directories, or a call-graph pass) that the brief author must provide in the row itself.
A passing verdict from this agent means "every citation is real and accurate." It does not mean "every conclusion is correct." The /research skill's Evidence Standard (code-symbol verdicts require grep evidence) is the author's responsibility; this agent does not backstop it.
If a brief row's verdict matches /remove|dead|unused|transitional|safe to (drop|remove)/i on a code symbol and the row does not cite grep-check liveness evidence, flag it under "Issues Found" as SEMANTIC-EVIDENCE-MISSING — but do not attempt the grep yourself (that is the author's obligation, not the verifier's).
Process
- Read the research brief provided to you
- Extract all citations into a verification queue:
- File paths:
path:linereferences to codebase files - URLs: External links to documentation, repos, or articles
- Document references: Named documents (ADRs, specs, guides)
- File paths:
- Verify each citation:
- File paths: Use Glob to confirm the file exists, then Read to confirm the line number is accurate
- URLs: Use WebFetch to confirm the URL resolves (look for 200 status or valid content)
- Document references: Use Grep to locate the document in the codebase
- Produce a verification report
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.
- 5d ago First seen · 72 lines · 75 tokens per session scan A e04b1602b51e
research-verifier is an agent published in the GitHub repository pitimon/8-habit-ai-dev (3 stars, last pushed 1mo ago), licensed MIT. It adds 75 tokens to every session and 817 once invoked, about $0.0004 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-reviewer
Review PR diffs for code quality, bugs, and style issues. Auto-triggered on PR creation. Not for: implementation, file editing, test writing.
challenger-calibration
Severity scoring guide with few-shot examples and hard thresholds for the challenger agent.
archiver
Archive project patterns and knowledge when retiring. Categorize as transferable, historical, or deletable. Use when retiring a project or archiving knowledge. Not for: active project work, pattern recall, daily operations.
memory-curator
Recall patterns from knowledge system, cross-pollinate across clients, check knowledge health. Use when starting work on a topic or periodic checks. Not for: codebase search, project scanning.
scout
Fast research for codebase exploration and web search. Use when finding files, searching code, or fetching external info. Not for: knowledge recall, project diagnostics, business analysis.
client-scanner
Scan client project structure, detect tech stack, check brana alignment. Use when entering an unfamiliar client project or for project health checks. Not for: business stage classification, knowledge recall, web research.