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 skills/progiri/craftlight/craft-graphnpx skills add progiri/craftlight --skill craft-graphgit clone --depth 1 https://github.com/progiri/craftlightWrote 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/progiri/craftlight/craft-graph)<a href="https://agentmods.dev/skills/progiri/craftlight/craft-graph"><img src="https://agentmods.dev/badge/skills/progiri/craftlight/craft-graph.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 | $0.00239 | $0.01930 |
| Opus 5 | $0.00120 | $0.00965 |
| Sonnet 5 | $0.00048 | $0.00386 |
| Haiku 4.5 | $0.00024 | $0.00193 |
Grade A, and why
craft-graph 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.
How it starts
The opening of the file, as written. The whole thing — 51 lines — stays where its author put it; the contents beside it link to each section on GitHub.
craft-graph — the graph of decisions and gotchas
Three principles on top of the shared ones: (1) the graph records what the code doesn't say on its own — "why we decided it this way" and "where the trap is", not a retelling of what's already visible; (2) every node is proven by a file:line proof — without proof it's guesswork, there's no node; (3) the graph is dual-view for free: the docs/graph/*.md nodes read as ordinary text, the [[wikilinks]] give a live graph in Obsidian, and a single _overview.md with Mermaid renders right on GitHub. Zero code edits and zero runtime — in the plugin's DNA.
Not for this skill: "how does this work", business-logic onboarding → codebase-reverse-engineering; "how to build/run/commands/conventions" → CLAUDE.md; review without edits → code-review; working out a decision not yet in the code → brief; changing code → task. The difference from code-wiki: craft-graph is a navigable graph with typed edges, an overview Mermaid, and a mandatory proof, not linear notes. Realized the task isn't about recording decisions as a graph → leave the skill.
Step 0. The object
Determine with cheap recon what we're documenting: the whole codebase / one subsystem / a specific module. Recon is targeted Grep/Glob and reading the fragments where decisions live — the canonical checklist is error handling, concurrency, caching, transaction boundaries, data invariants, working around non-obvious constraints (the modes point here); don't read whole files. The object isn't named and isn't obvious → one concrete question to the user.
docs/graph/already exists → an addition or refresh, not an empty start: check against existing nodes, extend them, don't breed duplicates. Nodes markedverifyin scope are the first candidates of a refresh pass — recheck the proof, update thefile:lineor split the node, clear the mark.docs/crafts/_backlog.mdhasgraph-candidatelines (left by S tasks / debug — a durable lesson with afile:lineinside) → these are ready-made node candidates; promote the worthy ones. Consume the breadcrumb: a promoted line is replaced in the backlog with a pointer→ [[slug]], not left as-is — an unconsumed breadcrumb re-promotes on the next pass and breeds duplicate nodes.- Resume of an interrupted Full is derivable, not stored: node areas are 1:1 with subsystems, so a subsystem with no node yet is either unvisited or decision-poor — check it and ask, don't restart the whole sweep.
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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 · 51 lines · 239 tokens per session scan A 61eed2430f48
craft-graph is a skill published in the GitHub repository progiri/craftlight (5 stars, last pushed 10d ago), licensed MIT. It adds 239 tokens to every session and 1,930 once invoked, about $0.0012 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 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
auto-perf-optimize
Run agent-driven VS Code performance or memory investigations. Use when asked to launch Code OSS, automate a VS Code scenario, run the Chat memory smoke runner, capture renderer heap snapshots, take workflow screenshots, compare run summaries, or drive a repeatable scenario before heap-snapshot analysis.
chat-perf
Run chat perf benchmarks and memory leak checks against the local dev build or any published VS Code version. Use when investigating chat rendering regressions, validating perf-sensitive changes to chat UI, or checking for memory leaks in the chat response pipeline.
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…