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 ayeshakhalid192007-dev/graph-engineering-crash-course --skill check-claimgit clone --depth 1 https://github.com/ayeshakhalid192007-dev/graph-engineering-crash-courseWrote 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/ayeshakhalid192007-dev/graph-engineering-crash-course/check-claim)<a href="https://agentmods.dev/skills/ayeshakhalid192007-dev/graph-engineering-crash-course/check-claim"><img src="https://agentmods.dev/badge/skills/ayeshakhalid192007-dev/graph-engineering-crash-course/check-claim/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/ayeshakhalid192007-dev/graph-engineering-crash-course/check-claim"><img src="https://agentmods.dev/badge/skills/ayeshakhalid192007-dev/graph-engineering-crash-course/check-claim.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 | $0.00055 | $0.01086 |
| Opus 5 | $0.00028 | $0.00543 |
| Sonnet 5 | $0.00011 | $0.00217 |
| Haiku 4.5 | $0.00006 | $0.00109 |
Grade A, and why
check-claim 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 9d 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 — 93 lines — stays where its author put it; the contents beside it link to each section on GitHub.
check-claim
This skill converts a claim shaped like "this change doesn't touch that
module" into a (change, touches, module) triple and searches the graph's
edge list for that exact subject/predicate/object combination. The
change's own description is not part of the input at any step: reading it
plays no role in whether the search finds a match, and no role in what
verdict that match produces.
Instructions
You are a Claude Code skill implementing the grounded-triple-checker
pattern. Follow these steps in order:
- Read the full graph first. Open
sample-graph.example.json(or the file the user names) and parse itsnodesandedgesarrays before doing anything else. - Resolve which claim(s) to check. If the user names a specific claim
id, check only that one. Otherwise default to every
claim-typed node in the graph — this kit's shipped scenario has two:claim-ch3002-no-license-touchandclaim-ch3047-no-license-touch. - Build the search triple. Pull the claim node's own
about(the change id) andforbidden_modulefields — do not parse these back out of the claim's free-texttextfield, and do not open the change node'sdescriptionat this step. Assemble them into(about, "touches", forbidden_module); step 4 walks the graph's edges looking for that combination. - Scan every edge in the graph, looking for one whose subject,
predicate, and object all three line up with that triple. A
touchesedge from the right change to a different module doesn't count, and neither does an edge of a different predicate between the same two nodes. - Decide the verdict from presence alone. A match found among the
graph's edges means REJECT, with that exact edge cited as the reason;
no match anywhere in the graph means ACCEPT. Do not consult the
change's
descriptionfield to double-check either outcome — a description that sounds convincing is not evidence, and neither is one that sounds suspicious. - Handle the unverifiable case explicitly. If the claim's
aboutchange id orforbidden_modulemodule id isn't found among the graph's nodes at all, report that claim as UNVERIFIABLE rather than defaulting it to ACCEPT. An edge that can't be looked up is not the same thing as an edge that's confirmed absent. - Write the results to
verdicts.json(or the path the user requested) in this kit's root: an array with one entry per claim, each holdingclaim_id,about,forbidden_module,decomposed_edge,edge_found,verdict, andcitation(the exact matching edge for a REJECT, ornullfor ACCEPT/UNVERIFIABLE). - Report a summary alongside the file: how many claims came back ACCEPT, REJECT, or UNVERIFIABLE, and the citation for every REJECT.
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.
- 9d ago First seen · 93 lines · 55 tokens per session scan A 3693fb75ac1b
check-claim is a skill published in the GitHub repository ayeshakhalid192007-dev/graph-engineering-crash-course (5 stars, last pushed 14d ago), licensed MIT. It adds 55 tokens to every session and 1,086 once invoked, about $0.0003 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
animate
Build a web animation: whether it should animate at all, the purpose, the tool, the properties, the easing curve, the duration, interruption and exit. Writes the CSS or Motion code for dropdowns, modals, toasts, tooltips, accordions, press feedback, stagger and scroll reveals. Use to animate something or make a…
building-agents
Use when building or restructuring an LLM agent — provider adapter, tool calling, structured output, RAG, agent loop, eval gate, cost routing, tracing, MCP server — model-agnostic across OpenAI/Anthropic/Gemini/OSS so a model swap is a config change. NOT vector-store SQL alone (that is postgresdb) or service…
postgresdb
Use when PostgreSQL engine behaviour decides the answer — schema and type design, index choice, reading EXPLAIN on a slow query, zero-downtime DDL and backfills, or ops (roles, RLS, pooling, vacuum, partitioning, PITR). PG16, ORM-agnostic. NOT portable query logic (that is sql), NOT a managed provider's platform…
author-skill
Use when authoring a NEW rsc skill or editing an existing one — scoping it to one job, writing the description that decides whether it ever loads, splitting the body into references/, writing its evals, auditing it against the rubric. NOT building a product feature (that is specify) and NOT designing an agent loop…
community
Use when building or running a persistent two-way community space — Discord, Telegram, Circle: platform choice, structure, onboarding, native→bot→human moderation, rituals, growth loops, health metrics. NOT churn of paying product customers (that is retention), NOT broadcast email (that is newsletter), NOT one live…
cpp
Use when writing, reviewing, modernizing, building, or debugging C++ - RAII and resource lifetime, smart-pointer ownership, move semantics and the Rule of Zero/Five, target-based CMake with FetchContent, and killing undefined behavior with ASan/UBSan/TSan plus clang-tidy. NOT borrow-checker / Result-Option / cargo…