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 three-layer-extension-syncgit 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/three-layer-extension-sync)<a href="https://agentmods.dev/skills/ychampion/cskill-agents/three-layer-extension-sync"><img src="https://agentmods.dev/badge/skills/ychampion/cskill-agents/three-layer-extension-sync/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/three-layer-extension-sync"><img src="https://agentmods.dev/badge/skills/ychampion/cskill-agents/three-layer-extension-sync.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 | $0.00027 | $0.00486 |
| Opus 5 | $0.00014 | $0.00243 |
| Sonnet 5 | $0.00005 | $0.00097 |
| Haiku 4.5 | $0.00003 | $0.00049 |
Grade A, and why
three-layer-extension-sync 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 — 30 lines — stays where its author put it; the contents beside it link to each section on GitHub.
SKILL: Three-Layer Extension Sync
Domain: plugin-installation Trigger: Apply when extensions exist simultaneously as declared settings, on-disk installs, and live runtime components, and those layers must stay coherent. Source Pattern: Distilled from reviewed plugin, extension, and MCP synchronization implementations.
Core Method
Split extension management into three layers: intent in configuration, materialization on disk, and active runtime state. Use the diff/reconcile phase to move layer 1 toward layer 2 without touching UI state, then use a separate refresh phase to rebuild layer 3 from the materialized result. This keeps each transition explicit, makes fallback rules clear, and prevents partial reload logic from smearing together configuration, installation, and activation responsibilities.
Key Rules
- Give each layer a single owner: configuration declares desired state, reconciliation mutates on-disk artifacts, and runtime refresh repopulates active commands, agents, hooks, and servers.
- Compare intent to materialization before doing work so you know whether to install, update, skip, or defer.
- Treat reconciliation as additive and side-effect-bounded; it should not directly mutate live AppState beyond progress or recovery flags.
- Refresh the active layer from cleared caches after materialization changes so runtime loaders rebuild from the new disk state instead of patching stale objects in place.
- Make escalation explicit: urgent changes can auto-refresh, while less urgent ones can stop at a needs refresh contract.
Example Application
In a desktop agent with third-party extensions, treat settings as the declared catalog, a local extensions directory as the materialized layer, and the in-memory command registry as the active layer. Reconcile settings to disk first, then run a distinct reload step that repopulates the active registry from the updated install set.
Anti-Patterns (What NOT to do)
- Do not let a single function both clone marketplaces and directly mutate live component registries; that blurs error handling and recovery.
- Do not compare runtime state back to settings as a proxy for disk state; materialization needs its own truth source.
- Do not patch active components incrementally against stale caches when the on-disk layer changed underneath them.
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 · 30 lines · 27 tokens per session scan A 23df9c2852b8
three-layer-extension-sync is a skill published in the GitHub repository ychampion/cskill-agents (36 stars, last pushed 5mo ago), licensed MIT. It adds 27 tokens to every session and 486 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 2026-09-03.
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.
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.
fix
Dispatch fix subagent for FIX-FIRST gaps from review, re-review, then diagnose unresolved failures after 2 loops.
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.