reelier-write-safety

reelier-write-safety is a skill for Codex from seldonframe/reelier. It costs 78 tokens per session (3,263 once invoked), scanned B, original, MIT.

A safety guide for controlling and recording what an AI agent writes to real systems. It uses Reelier, a command-line tool for checking write coverage, approvals, policies and replay records.

In plain words
What is it for?
Use it before giving an agent write access to systems such as a CRM, booking tool, advertising account or knowledge store.
Why use it?
It reduces the risk of an unattended or scheduled agent changing important data without clear limits or an audit trail.

Skill for Codex

Written for Codex: reads ~/.codex or $CODEX_HOME. Also seen: reads .claude/ paths; mentions Claude Code; mentions Codex.

Good fit Use it before giving an agent write access to systems such as a CRM, booking tool, advertising account or knowledge store.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/seldonframe/reelier/reelier-write-safety
Install

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.

Any agent
npx skills add seldonframe/reelier --skill reelier-write-safety
Clone the repo
git clone --depth 1 https://github.com/seldonframe/reelier

Made for: Codex.

Wrote 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.

agentmods badge for reelier-write-safety

README.md
[![agentmods](https://agentmods.dev/badge/skills/seldonframe/reelier/reelier-write-safety/github.svg)](https://agentmods.dev/skills/seldonframe/reelier/reelier-write-safety)
Your own site
<a href="https://agentmods.dev/skills/seldonframe/reelier/reelier-write-safety"><img src="https://agentmods.dev/badge/skills/seldonframe/reelier/reelier-write-safety/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for reelier-write-safety

Your own site · 80×15
<a href="https://agentmods.dev/skills/seldonframe/reelier/reelier-write-safety"><img src="https://agentmods.dev/badge/skills/seldonframe/reelier/reelier-write-safety.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 78 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,263 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00078 $0.03263
Opus 5 $0.00039 $0.01631
Sonnet 5 $0.00016 $0.00653
Haiku 4.5 $0.00008 $0.00326

Measured 9d ago against content hash f7f1924827da, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade B, and why

reelier-write-safety scanned grade B 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 9d 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.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

`~/.codex/config.toml` — it is TOML, and the installer only parses JSON. A
integrations/skills/reelier-write-safety/SKILL.md · 263 lines

How it starts

The opening of the file, as written. The whole thing — 263 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Reelier — bound the write before you grant it

written against reelier 0.32.x. reelier --help on the installed CLI is always authoritative.

If nothing consequential gets written, stop here

If the tool being wired up only reads — a docs lookup, a search index, a read-only status API — say so in one sentence and move on. Bounding a read is ceremony nobody asked for. The rest of this skill is for the moment an agent gets to change something real: a CRM row, a booking, an ad budget, a knowledge store, a branch.

How to run Reelier — in this order

  1. The reelier_* MCP tools do not help here. reelier serve exposes only reelier_scan, reelier_from_session, reelier_replay, reelier_push and reelier_diff — the record-and-replay half. None of coverage, install, policy check or approve is reachable over MCP, so every command in this skill goes through the CLI.
  2. Run the CLI: npx -y reelier <command>. Nothing needs to be installed first.
  3. If there is no shell — no way to run npx — say plainly that the CLI is unavailable and stop. Never describe a step as done when it did not run.

Two Reelier commands sound alike and do opposite things. Keep them apart or every sentence you write about coverage will be wrong:

  • reelier serve is a tool-server: it exposes Reelier's own commands (scan, from-session, replay, push, diff) to your agent as MCP tools. It observes nothing and wraps nothing.
  • reelier mcp --wrap "<command>" is the recorder: it starts another MCP server as a child process and sits in front of it, so every call that server receives is recorded. This is the one that produces coverage.

Adding serve to the user's project MCP config gives an agent the record-and-replay tools without shelling out. It does not put a recorder in front of anything, and it does not expose any of the commands below.

What Reelier bounds — and what it flatly refuses to

Reelier bounds scope and change: which tools may be called, whether a write was approved before it fired, and what the system looked like after compared with what was declared.

Read the full file on GitHub · 263 lines

Changes

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.

  1. 9d ago First seen · 263 lines · 78 tokens per session scan B f7f1924827da

Subscribe to this mod's changes

reelier-write-safety is a skill published in the GitHub repository seldonframe/reelier (3 stars, last pushed 15d ago), licensed MIT. It adds 78 tokens to every session and 3,263 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

robin

Use whenever an agent creates, updates, reviews, or completes a GitHub pull request. Detect whether Robin is installed in the repository; when it is, automatically drive the PR through a bounded review, verified-fix, reply, thread-resolution, re-review, authorized-merge, and cleanup loop without requiring the user to…

antongulin/robin · 92 tokens

shell-scripting

Use this skill when writing bash or zsh scripts, parsing arguments, handling errors, or automating CLI workflows. Triggers on bash scripting, shell scripts, argument parsing, process substitution, here documents, signal trapping, exit codes, and any task requiring portable shell script development.

alibaba/anolisa · 60 tokens

ai-unity-asset-pipeline

Build, install, run, and verify a reusable image-to-game-ready-3D pipeline for Unity with controlled ImageGen references, approved AI mesh providers, Blender MCP/CLI, modular era catalogs, Blender retopology/LOD/pivot processing, and Unity import/render proof. Use for 2D-to-3D assets, vehicles, aircraft, ships…

ezBuilder/code-brain · 95 tokens

memory-lint

Deterministic integrity check for the agent-memory layer. Use after a memory review, before committing memory/ changes, or in CI to catch decay miscounts — facts archived while still referenced, an id in both continuity and the archive, tier or supersession drift — and leaked secrets or PII (credentials, tokens…

acn-ericlaw/agent-memory · 89 tokens

second-opinion

Generate a fresh-context review snapshot of the current task for a clean-memory AI reviewer (any vendor or a clean session). Use at a milestone that looks complete, or when blocked, uncertain, or facing a risky change, to get an independent second opinion before proceeding; shows a security advisory before exporting…

acn-ericlaw/agent-memory · 66 tokens

archive-fact

Safely move a decayed fact or completed thread from continuity.md into the archive — the mechanical half of a memory review, done deterministically. Use during the REVIEW.md ritual (step 4) or whenever a specific fact must be archived, instead of hand-editing continuity.md (which has truncated the file twice). The…

acn-ericlaw/agent-memory · 81 tokens