Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/kanetran29/uigraphnpx agentmods add skills/kanetran29/uigraph/gateWrote 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/kanetran29/uigraph/gate)<a href="https://agentmods.dev/skills/kanetran29/uigraph/gate"><img src="https://agentmods.dev/badge/skills/kanetran29/uigraph/gate/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/kanetran29/uigraph/gate"><img src="https://agentmods.dev/badge/skills/kanetran29/uigraph/gate.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 | $0.00075 | $0.00782 |
| Opus 5.5 | $0.00030 | $0.00313 |
| Sonnet 5.5 | $0.00015 | $0.00156 |
| Haiku 4.5 | $0.00007 | $0.00078 |
Grade A, and why
uigraph-gate 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 3d 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 — 60 lines — stays where its author put it; the contents beside it link to each section on GitHub.
uigraph verification gate
The gate (.uigraph/kit/gate.mjs) is the machine verifier that proves a UI
change matches its ticket's acceptance criteria. Six verdict layers, fail-fast:
check-cmd → ast-grep → impact containment → graph-match → verify → specs.
Invocation
node .uigraph/kit/gate.mjs baseline <app-dir>
node .uigraph/kit/gate.mjs check <app-dir> --contract <contract.json> \
[--no-browser | --app-url http://localhost:PORT]
Prerequisites: Node >=22, uigraph on PATH (or set UIGRAPH_BIN), ast-grep
on PATH, @playwright/test as a devDependency for the behavioral layer
(structure layers run without it). --check-cmd "<cmd>" overrides verdict 1
(default pnpm check).
Exit: 0 GREEN · 1 an AC failed (read <app-dir>/.uigraph-gate/report.md)
· 2 infrastructure problem — fix the environment; does NOT count as a
failing round.
Protocol
- Baseline before touching code:
gate.mjs baseline <app-dir>(pins the pre-code graph + git sha), once per ticket. - Translate the ACs → a contract (
contracts/<ticket>.json). Smoke-test every ast-grep pattern against a should-match scratch fixture first (forbid-mode fails open on a typo). STOP and get the human's approval of the contract — approval freezes the definition of done. You translate; you never grade. - Code loop: implement →
check→ on exit 1 read the report and fix. Max 3 rounds; the same failing-verdict set 2 rounds in a row → stop early, escalate. Use--no-browserfor fast mid-loop runs; the final proving run MUST use--app-urlwith the dev server up. - GREEN → report to the human with the report file. Never claim a feature done on a red or skipped browser verdict.
Watch-outs
- forbid-mode patterns fail OPEN on a typo. ast-grep silently matches
nothing for a malformed pattern, so a
mode: "forbid"rule with a typo never fires and the verdict stays green. Smoke-test each forbid pattern once against a scratch file that SHOULD match before trusting it. - ast-grep cannot see inside
.vueSFCs. It skips.vuefiles while walking a directory, so aforbid-mode pattern goes red instead of silently trusting a clean 0-match run when the workspace has.vuefiles. Opt a pattern out — only when certain it concerns non-.vuecode — with the[ts-only]tag:"pattern[ts-only]: setId($_)". verifyreds on pre-existing graph damage, not just this ticket. An unrelated unverified/refutable transition elsewhere in the graph can red the gate even though your change is correct. Keep the workspace's graph healthy rather than treating every redverifyas your ticket's fault.
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.
- 3d ago First seen · 60 lines · 75 tokens per session scan A b618e57aa7f1
uigraph-gate is a skill published in the GitHub repository kanetran29/uigraph (5 stars, last pushed 3d ago), licensed MIT. It adds 75 tokens to every session and 782 once invoked, about $0.0003 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 2026-11-04.
uigraph-gate compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| uigraph-gatekanetran29/uigraph | Skill | 5 | 3d ago | MIT | A | 75 | Claude Code, Codex |
| weavatrixWeavatrix/weavatrix | Skill | 4 | 1mo ago | MIT | A | 108 | Cursor |
| weavatrixWeavatrix/weavatrix | Skill | 4 | 1mo ago | MIT | A | 108 | Codex, Cursor |
| opik-diagnosecomet-ml/opik-mcp | Skill | 225 | 9d ago | Apache-2.0 | A | 187 | Claude Code |
| dcc-blender-assembly-pitfallsdcc-mcp/dcc-mcp-blender | Skill | 60 | yesterday | MIT | A | 153 | Claude Code |
| virality-analysismtarcure/claude-vibe-squad | Skill | 228 | 9d ago | MIT | A | 56 | Claude Code, Codex |
| color-theorymtarcure/claude-vibe-squad | Skill | 228 | 9d ago | MIT | A | 62 | Claude Code, Codex |