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 transcript-to-claimsgit 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/transcript-to-claims)<a href="https://agentmods.dev/skills/ayeshakhalid192007-dev/graph-engineering-crash-course/transcript-to-claims"><img src="https://agentmods.dev/badge/skills/ayeshakhalid192007-dev/graph-engineering-crash-course/transcript-to-claims.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 5.1 | $0.00033 | $0.00896 |
| Opus 5 | $0.00016 | $0.00448 |
| Sonnet 5 | $0.00007 | $0.00179 |
| Haiku 4.5 | $0.00003 | $0.00090 |
Grade A, and why
transcript-to-claims 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 — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.
transcript-to-claims
Turns a conversation transcript into a set of provisional claim nodes — one per factual assertion — without resolving, merging, or deciding which claims are true. Hedged language stays marked as hedged.
Instructions
You are a Claude Code skill implementing the conversation-to-claims
pattern. Work through the transcript turn by turn, in order. For each
turn:
- Decide whether the turn contains an assertion at all. A question ("can you tell me...") or a pure request ("can you send someone out") contributes no claim node on its own. Only sentences that state something about the world count.
- Split multi-assertion turns. A single turn can carry more than one distinct factual statement (for example, one sentence about a repair date and a separate sentence about when a symptom started). Give each distinct assertion its own claim node rather than folding a turn's entire content into one.
- Extract subject/predicate/object where the sentence supports it. If the sentence doesn't cleanly reduce to a triple, keep the claim's raw text instead of forcing a bad fit — a claim node with accurate free text beats one with a fabricated triple.
- Check for hedge language before finalizing each claim. Words and
phrases like "might," "I think," "possibly," "not certain," "as far
as I know" mark the assertion as the speaker's belief rather than a
stated fact. Set a
hedged: trueflag on that claim node and keep the hedge language visible in its text — do not strip it out or round the claim up to a flat assertion. - Do not resolve, merge, or deduplicate claims here. Even if two claims from different speakers appear to overlap or conflict, write both out as separate provisional nodes and note the apparent overlap in your report — resolving it is a later stage's job, not this skill's.
- Attach speaker and turn number to every claim node, so a later reader (or the resolution stage) can trace each claim back to exactly where in the transcript it came from.
- Report the full set of claim nodes in turn order, each labeled with speaker, turn number, hedged/not-hedged, and the claim itself. State explicitly which turns produced no claims and why (question, request, small talk).
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 · 86 lines · 33 tokens per session scan A a5b8f656f98e
transcript-to-claims is a skill published in the GitHub repository ayeshakhalid192007-dev/graph-engineering-crash-course (5 stars, last pushed 13d ago), licensed MIT. It adds 33 tokens to every session and 896 once invoked, about $0.0002 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
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…
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…
gamedev-multiplayer
Use when adding multiplayer or netcode to a game — client-server vs P2P, server authority and anti-cheat, state replication vs RPCs, prediction and reconciliation, lag compensation, plus Godot 4 / Unity NGO / Unreal wiring. NOT single-player gameplay (that is godot, unity, unreal), NOT matchmaking or server hosting…
gamedev-physics
Use when working with a game engine's physics — rigid bodies, colliders, collision layers and masks, character controllers, joints, forces versus impulses, raycasts and shapecasts, or when a simulation is unstable (tunnelling, jitter, missed overlaps). Covers the Godot, Unity and Unreal equivalents. NOT rendering or…