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 instructions/avansaber/tailtest/claude-mdgit clone --depth 1 https://github.com/avansaber/tailtestWrote 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/instructions/avansaber/tailtest/claude-md)<a href="https://agentmods.dev/instructions/avansaber/tailtest/claude-md"><img src="https://agentmods.dev/badge/instructions/avansaber/tailtest/claude-md.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 | $0.07383 | $0.07383 |
| Opus 5 | $0.03691 | $0.03691 |
| Sonnet 5 | $0.01477 | $0.01477 |
| Haiku 4.5 | $0.00738 | $0.00738 |
Grade A, and why
tailtest CLAUDE.md scanned grade A with 1 finding 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 3d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
Not just `requests` / `axios`. Include `smtplib`, `socket`, `urllib`, `http.client`, `ftplib`, `imaplib`, and any `subprocess` call reaching an external process. How it starts
The opening of the file, as written. The whole thing — 426 lines — stays where its author put it; the contents beside it link to each section on GitHub.
tailtest
You are running with the tailtest plugin. Your job: automatically run the test cycle the user would otherwise ask for manually. Generate production-like scenarios for what was just built, execute them, and surface only what fails.
Step 0: Verify APIs before writing test code
Before writing any test code, read the source file and confirm:
- Every class, function, and method you plan to call in the test actually exists in the module
- Every import resolves to a real name in the file
Scope: verifies imports resolve and named attributes exist. Does not validate full call signatures. If a method does not exist, do not invent it -- adjust the scenario to test what actually exists.
Step 1: At the start of every user turn, check for pending work
Read .tailtest/session.json. If pending_files is non-empty:
Before generating: re-read the source file to understand what it actually does. Derive scenarios from the source's intent and behaviour -- not from your implementation plan or assumptions about what should exist.
- Note the pending files list
- Skip any filtered files (filter rules in Step 2)
- If nothing remains after filtering: clear
pending_filesto[], proceed to the user's message - Before generating, check for ramp-up framing: if
session.jsonhas"ramp_up": true, count the entries whosestatusis"ramp-up"(call this N). If N > 0 and all remaining pending_files entries havestatus: "ramp-up"(pure ramp-up batch, no mixed new-file entries), emit exactly one line now:tailtest: running initial coverage scan on {N} file(s)...If the batch is mixed (contains both"ramp-up"and"new-file"entries), skip the framing line. - Generate scenarios for all remaining files as one unit of work -- treat
"ramp-up"entries identically to"new-file"(generate scenarios, write test, execute, report) - Write test file and execute (Steps 4–5)
- Report failures -- stay silent if all pass (Step 6)
- Write
"pending_files": []back to.tailtest/session.json - Then address the user's 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.
- 3d ago First seen · 426 lines · 7,383 tokens per session scan A ec88608cff60
tailtest CLAUDE.md is an instructions file published in the GitHub repository avansaber/tailtest (11 stars, last pushed 2mo ago), licensed MIT. It adds 7,383 tokens to every session, about $0.0369 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
plumbline CLAUDE.md
Instructions for sroberts/plumbline, covering claude.md, project status, build & test, architecture in brief and generated and derived files — don't hand-edit.
agentic-tech-debt AGENTS.md
Instructions for bcanfield/agentic-tech-debt: The agent instructions for this repo live in CLAUDE.md — one source of truth for every coding agent. Read it and follow it as written; it applies to Codex exactly as it does to Claude Code.
zuvo CLAUDE.md
Instructions for greglas75/zuvo, covering zuvo plugin — project guide, what this repo is, tech stack, how to update after making changes and for yourself (dev testing, no marketplace push).
ai-pr-bot CLAUDE.md
Instructions for sgharlow/ai-pr-bot, covering claude.md, common development commands, install dependencies, development mode with hot reload and build the project.
agentic-hardening-skill AGENTS.md
Instructions for Calvin-LLC/agentic-hardening-skill, covering harden and rules.
repomemory CLAUDE.md
Claude Code instructions for PraveenJayaprakash-JP/repomemory, covering claude.md — repomemory, project overview, architecture, commands and root (next.js web app).