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/seldonframe/reelier/claude-mdgit clone --depth 1 https://github.com/seldonframe/reelierWrote 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/seldonframe/reelier/claude-md)<a href="https://agentmods.dev/instructions/seldonframe/reelier/claude-md"><img src="https://agentmods.dev/badge/instructions/seldonframe/reelier/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 | $0.06935 | $0.06935 |
| Opus 5 | $0.03467 | $0.03467 |
| Sonnet 5 | $0.01387 | $0.01387 |
| Haiku 4.5 | $0.00694 | $0.00694 |
Grade A, and why
reelier CLAUDE.md 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 4d 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.
This is a copy
100% identical to reelier AGENTS.md — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 383 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reelier — what it actually does
Release snapshot: [email protected], verified 2026-08-10. npm serves v0.32.0. The published
artifact predates the certification additions now on codex/outcomes-delegation-infra; the Cloud branch is
codex/outcomes-cloud. Production Cloud is deployed and database migrations are applied, but live
provider certification and the real ten-agent Codex run remain gated on isolated resources and
credentials. Every claim below must be re-verified against code and live evidence before saying a
capability is production-certified.
The branch target is 0.32.1, not a republish of 0.32.0. Cloud must be pinned and retested
against the exact final 0.32.1 tarball before either PR merges.
What the 2026-08-06 pass found, as a warning about this file's own failure mode. The previous pin (
62cd841, 2026-08-01) claimed policy attestation was "fixed on main, NOT yet released" and that npm still carried the silent version. By 2026-08-06npm view reelier versionreturned 0.30.0 — it had shipped. The file spent an unknown number of days asserting a false and pessimistic claim about a safety property. Two commands would have caught it. Run §10.1.
Why this file is pinned. A capabilities doc is the exact artifact that silently goes stale — the failure class this product exists to catch. An unpinned version of this file would be a guardrail that is present, silent, and dead.
1. One-line identity
Mission: "Let agents write." Every autonomous write bounded before it happens, attested after it happens, answerable forever. Reelier sells capability ("this is how you let your agents write"), never surveillance ("this is how you watch what they did").
The bet: write-trust is a state-transition problem, not a model problem. You make a write grantable by bounding and proving what it changes, not by evaluating what the agent thinks.
2. THE THING MOST OFTEN GOT WRONG — there are two execution paths
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.
- 4d ago First seen · 383 lines · 6,935 tokens per session scan A 4407474a3809
reelier CLAUDE.md is an instructions file published in the GitHub repository seldonframe/reelier (3 stars, last pushed 10d ago), licensed MIT. It adds 6,935 tokens to every session, about $0.0347 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to reelier AGENTS.md, differing in 0 lines, and is treated as a copy.
Other instructions, from other repositories
agent-memory CLAUDE.md
Claude Code instructions for acn-ericlaw/agent-memory: Both a shared AI memory layer and an AI-enablement / migration tool — no-code, markdown.
agent-memory copilot-instructions.md
Copilot instructions for acn-ericlaw/agent-memory: Both a shared AI memory layer and an AI-enablement / migration tool — no-code, markdown.
agent-memory AGENTS.md
AGENTS.md instructions for acn-ericlaw/agent-memory, a project described as: A lightweight, vendor-neutral AI memory tool for human–AI collaboration that turns a clear, shared vision into deterministic delivery with near-zero drift.
agent-memory GEMINI.md
Gemini CLI instructions for acn-ericlaw/agent-memory: Both a shared AI memory layer and an AI-enablement / migration tool — no-code, markdown.
m_flow AGENTS.md
AGENTS.md instructions for FlowElement-xinliuyuansu/m_flow, covering m-flow — developer & agent reference, 1. repository map, extension points, 2. local development and python backend (requires python 3.10 – 3.13).
html-anything AGENTS.md
AGENTS.md instructions for nexu-io/html-anything, covering this is not the next.js you know, workspace shape and commands.