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 GizClaw/flowcraft --skill flowcraft-configgit clone --depth 1 https://github.com/GizClaw/flowcraftWrote 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/gizclaw/flowcraft/flowcraft-config)<a href="https://agentmods.dev/skills/gizclaw/flowcraft/flowcraft-config"><img src="https://agentmods.dev/badge/skills/gizclaw/flowcraft/flowcraft-config/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/gizclaw/flowcraft/flowcraft-config"><img src="https://agentmods.dev/badge/skills/gizclaw/flowcraft/flowcraft-config.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Anti-Refusal · line 64 Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.Fix: Remove instructions that suppress warnings, disclaimers, or ethical commentary. Let the agent surface safety-relevant caveats to the user.
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 | $0.00121 | $0.02089 |
| Opus 5.5 | $0.00048 | $0.00836 |
| Sonnet 5.5 | $0.00024 | $0.00418 |
| Haiku 4.5 | $0.00012 | $0.00209 |
Grade A, and why
flowcraft-config 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 today.
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 — 159 lines — stays where its author put it; the contents beside it link to each section on GitHub.
FlowCraft Config Authoring
Write complete FlowCraft deployment configuration: deploy.yaml, the
runtime section, resource sub-documents, and graph JSON. Validate with
the L2 structural validator and fix errors against the reference cards.
Workflow
- Scope the deployment. Collect the agents, capabilities (chat, tools, memory, scripts, sandboxing), and runtime needs (sessions, resume, event bus, checkpoint stores).
- Draft the deployment document.
deploy.yamlis the convention but any filename works; pass whatever path you choose to the validator. Read references/deploy.md first. Order the areas: resources → agents → runtime. Decide whole-resource vs item dep refs (infervsws/project). - Write sub-documents. Read
references/resources.md for the owning
module's schema. Workspace before sandbox (custom sandbox runners may
depend on workspace resources). Memory implementation modules are
app-registered — omit implementation examples;
core/memorycontracts and hooks are fine to use. - Write graph JSON. Read references/graph.md.
Prefer routing: wire the
inference.Routerinto the graph engine and omitmodelin inference nodes; pinmodelonly when no router is wired. Model refs must use the nestedidform; script nodes needruntimeandsource; wire edges back to the inference node after tool nodes. - Assemble the Craft application. When the deployment runs inside a
Craft, wrap it in
craft.yaml: an inlinedeploy(orbase_layers), plusui,plugins(roots,tool_registry,node_targets) andhost_toolswhere they matter. Read references/craft.md, then copy assets/minimal-craft as a starting point. A definition is checked by the host withcraft.ParseDefinition; the standalone validator does not covercraft.yaml(see "Compatibility and versioning"). - Validate structurally with L2. Run
skills/flowcraft-config/scripts/validate-config.sh <deployment-file>(or the installed copy's script). The validator pins the FlowCraft core module in itsgo.modand works standalone from any directory. It is structural only: it strictly decodes the document throughcore/deploy.Parse(unknown fields are rejected, resource/agent entries are shape-checked), strictly decodes and validates theruntimesubtree throughcore/runtime.DecodeConfigwhen present, and structurally validates graph definitions throughcore/graph.GraphDefinition.Validate. It does not build resources: no factory registry, no settingsfile/embedresolution, no node config decoding, no credentials, and no provider calls. Custom and app-registered kinds pass as long as they fit the resource envelope; their settings semantics are the host build's job.
What ships with it
22 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.
- agents/openai.yaml 302 B
- assets/minimal-craft/craft.yaml 1.9 KB
- assets/minimal-craft/graphs/assistant.json 448 B
- assets/minimal-craft/inference.yaml 768 B
- assets/minimal-craft/plugins/hello/plugin.json 129 B
- assets/minimal-craft/plugins/hello/skills/hello/SKILL.md 484 B
- assets/minimal-craft/workspace.yaml 224 B
- assets/minimal-deploy/deploy.yaml 781 B
- assets/minimal-deploy/graphs/assistant.json 448 B
- assets/minimal-deploy/inference.yaml 1.2 KB
- assets/minimal-deploy/workspace.yaml 218 B
- assets/minimal-deploy/workspace/.gitkeep 1 B
- references/craft.md 18 KB
- references/deploy.md 5.8 KB
- references/graph.md 22 KB
- references/pitfalls.md 8.8 KB
- references/resources.md 44 KB
- references/runtime.md 4.9 KB
- scripts/validate-config.sh 645 B runs code
- scripts/validator/go.mod 1.9 KB
- scripts/validator/go.sum 8.8 KB
- scripts/validator/main.go 2.5 KB
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.
- today Changed a3f810f38a0d
- 11d ago Changed · +31 lines · +25 tokens per session d4bf6fcc8c97
- 14d ago Changed 7f462209a9cc
- 1mo ago Changed 9a709faf5c61
- 1mo ago Changed eeb341f2b8d3
- 1mo ago Changed bbab33133e7a
- 1mo ago First seen · 128 lines · 96 tokens per session scan A 6635f5b378d3
flowcraft-config is a skill published in the GitHub repository GizClaw/flowcraft (416 stars, last pushed yesterday), licensed MIT. It adds 121 tokens to every session and 2,089 once invoked, about $0.0005 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
vuln-scanner
Audit trending repos for real security vulnerabilities and disclose responsibly - scan and route findings (PVR / dependency PR), re-submit queued advisories, and send armed email disclosures.
token-movers
Crypto market scanner and single-token analyst - movers scans top winners/losers/trending or on-chain runners with pump-risk flags; single-token produces a verdict-first deep report for one token.
vuln-tracker
One lifecycle poll over everything vuln-scanner produces - PR and advisory status, PVR triage transitions, and pending-disclosure aging, with a stars-secured impact headline and one action queue.
aeon-update
Pull framework updates from the upstream Aeon repo into this instance - 3-way merges canon's new commits into a PR, never clobbering operator config.
article
Write a publication-ready article in one of three angles - a trending long-form piece, a watched-repo thesis, or a project-through-a-lens essay. Optional Replicate hero image with --visual.
distribute-tokens
Two-phase contributor rewards - plan builds a tier-priced payout from the repo's merged-PR ranking; send executes it on-chain via Bankr Wallet API with per-recipient idempotency and dry-run.