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 skills add ychampion/cskill-agents --skill declarative-read-only-command-allowlistsgit clone --depth 1 https://github.com/ychampion/cskill-agentsWrote 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/skills/ychampion/cskill-agents/declarative-read-only-command-allowlists)<a href="https://agentmods.dev/skills/ychampion/cskill-agents/declarative-read-only-command-allowlists"><img src="https://agentmods.dev/badge/skills/ychampion/cskill-agents/declarative-read-only-command-allowlists.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 | $0.00026 | $0.00448 |
| Opus 5 | $0.00013 | $0.00224 |
| Sonnet 5 | $0.00005 | $0.00090 |
| Haiku 4.5 | $0.00003 | $0.00045 |
Grade A, and why
declarative-read-only-command-allowlists 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 8d 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 — 28 lines — stays where its author put it; the contents beside it link to each section on GitHub.
SKILL: Declarative Read Only Command Allowlists
Domain: shell-validation Trigger: Use when you want tooling to enforce read-only shell commands by referencing a static map of command names, allowed flags, and supplemental callbacks. Source Pattern: Distilled from reviewed permission, shell-safety, and worktree-management implementations.
Core Method
Define each safe command as a declarative entry keyed by its full name such as git diff, docker logs, or rg. Each entry should describe allowed flags, argument rules, and any extra validation needed for edge cases. Build the shell policy layer by consulting these tables instead of scattering safety logic across the executor. The executor then only needs to look up the command, validate its flags against the declarative spec, and reject anything not explicitly allowed.
Key Rules
- Keep every command entry immutable so runtime validators can memoize them and share the same configuration between BashTool and PowerShellTool.
- Prefer breaking large maps (git, docker, rg) into shared flag groups (e.g., git stat flags, git color flags) to avoid duplication and to make audits easy.
- Allow supplemental callbacks only for the few commands that need extra context (e.g.,
git remote showneeds remote-name validation) and keep them side-effect-free. - Treat unknown commands or flags as rejects, making the declarative tables the only source of truth for allowed operations.
Example Application
When building a new shell agent, load the read-only command map, validate the tokenized flags for each request against that map, and allow execution only when the command matches an approved entry. This makes the policy auditable and consistent across shells.
Anti-Patterns (What NOT to do)
- Don’t hardwire flag lists inside the tool executor; that risks drift between
git diffand the security policy. - Don’t extend the tables casually; every new command or flag should be justified and validated against the same argument rules as the rest of the policy.
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.
- 8d ago First seen · 28 lines · 26 tokens per session scan A 6728988ba0c6
declarative-read-only-command-allowlists is a skill published in the GitHub repository ychampion/cskill-agents (36 stars, last pushed 5mo ago), licensed MIT. It adds 26 tokens to every session and 448 once invoked, about $0.0001 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 2026-08-30.
Other skills, from other repositories
grape
Use Grape MCP for Codex context continuity in coding repositories. Use when a task needs repeated-turn context, omitted context restore, stale-context checks, invalidation checks, or safe continuity across branch and dirty-worktree changes.
opensrc
Fetch dependency source code to give AI agents deeper implementation context. Use when the agent needs to understand how a library works internally, read source code for a package, fetch implementation details for a dependency, or explore how an npm/PyPI/crates.io package is built. Triggers include "fetch source for"…
review
Validate plans, execution, or PRs against wish criteria — returns SHIP / FIX-FIRST / BLOCKED with severity-tagged gaps.
work
Execute an approved wish plan — orchestrate subagents per task group with fix loops, validation, and review handoff.
brainstorm
Explore ambiguous or early-stage ideas interactively — tracks wish-readiness and crystallizes into a design for wish.
genie
Entry point for Genie operations — routes bug reports, questions, and operational commands, resumes existing lifecycle state, and orchestrates work that needs durable planning or coordination. Other ordinary requests bypass the lifecycle with a one-line notice unless the user asks for Genie.