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 fork-child-directive-guardrailsgit 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/fork-child-directive-guardrails)<a href="https://agentmods.dev/skills/ychampion/cskill-agents/fork-child-directive-guardrails"><img src="https://agentmods.dev/badge/skills/ychampion/cskill-agents/fork-child-directive-guardrails/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.
<a href="https://agentmods.dev/skills/ychampion/cskill-agents/fork-child-directive-guardrails"><img src="https://agentmods.dev/badge/skills/ychampion/cskill-agents/fork-child-directive-guardrails.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00024 | $0.00428 |
| Opus 5.5 | $0.00010 | $0.00171 |
| Sonnet 5.5 | $0.00005 | $0.00086 |
| Haiku 4.5 | $0.00002 | $0.00043 |
Grade A, and why
fork-child-directive-guardrails 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 1mo 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: Fork Child Directive Guardrails
Domain: agent-coordination Trigger: Apply this when a forked sub-agent inherits a conversation and must obey boilerplate guard text, cache-safe output, and explicit directive rules before running. Source Pattern: Distilled from reviewed subagent orchestration, isolation, and lifecycle implementations.
Core Method
Compose the fork prompt by merging the parent assistant message with placeholder tool results, then append a distilled directive that enumerates the guardrails (no sub-agents, no chatting, execute silently). Explicitly include the <fork-boilerplate> block and the fork directive prefix so downstream cache sharing sees identical bytes for repeated forks. The directive should reinforce pointer translation, note the isolated worktree, and remind the worker that system prompts do not apply.
Key Rules
- Preserve the parent assistant message intact before the fork; do not rewrite tool labels or thinking blocks so the prompt cache remains compatible.
- Build uniform placeholder tool_result blocks with fixed text (
Fork started — processing in background) and append the directive block that lists the 10 non-negotiable rules before any user-facing output. - Include the
<fork-boilerplate>tag per current cache strategy and apply fork directive prefix to keep child messages byte-identical across runs. - Explicitly mention that the worker is operating in a fork/isolated worktree and must translate inherited paths before editing.
Example Application
When imploding a background worker to handle tool-heavy tasks, inject this guardrail directive before executing the fork; downstream watchers will see the same prompt scaffold and track the worker as a fork without extra moderation.
Anti-Patterns (What NOT to do)
- Do not omit the
<fork-boilerplate>block or use variable placeholder text, which would invalidate cache sharing. - Do not let the fork spawn more agents or chat with humans — the directive must forbid sub-agent usage and conversation.
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.
- 1mo ago First seen · 28 lines · 24 tokens per session scan A 978a81a1a92d
fork-child-directive-guardrails is a skill published in the GitHub repository ychampion/cskill-agents (36 stars, last pushed 6mo ago), licensed MIT. It adds 24 tokens to every session and 428 once invoked, about $0.0001 per session on Opus 5.5. 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.
gh
GitHub CLI skill for interacting with GitHub via the gh command line tool. Use when Bub needs to (1) Create, view, or manage GitHub repositories, (2) Work with issues and pull requests, (3) Create and manage releases, (4) Run and monitor GitHub Actions workflows, (5) Create and manage gists, (6) Install and manage…
brain-ingest
The process for digesting a conversation, document, or research result, classifying it, and writing it down as brain content (a root-page update or a new/updated page) through the brain CLI.
wish
Deliver one decided task end to end — admit it, work it in one worktree, gate, independent review, bounded repair, a merge-ready PR — or plan a multi-group wish when it is bigger than one task.
gate-control
Canonical VideoStudio review authorization and state-transition policy. Use after any Gate B/C/Preview/D decision, post-gate revision, resumed approval, or exhausted visual-QA result across COMPOSE/AUTO/GENERATE/EDIT; maps explicit user authority and durable artifact state to one next action with ovs gate transition.…
report
Investigate a failure to its root cause with grounded evidence, hand the diagnosis to fix, and create a GitHub issue only when asked.