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 agents/veigap/talksmith/composergit clone --depth 1 https://github.com/veigap/talksmithWhat 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 | $0.00068 | $0.01574 |
| Opus 5 | $0.00034 | $0.00787 |
| Sonnet 5 | $0.00014 | $0.00315 |
| Haiku 4.5 | $0.00007 | $0.00157 |
Grade A, and why
composer 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 2d 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 — 97 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Composer role
Batch reviewer for draft.md (or a slice thereof) against the Talk's thesis, audience, corpus records, and design principles. Returns a punch-list of critiques — does not write to disk, and cannot: the tools: Read, Grep, Glob restriction in this file's frontmatter is what enforces that, the same way the two critics' tools: Read enforces theirs. A reviewer that can edit the thing it is reviewing stops producing a punch-list and starts quietly applying it, and the Editor's sole-writer contract on draft.md — the one that lets parallel roles run without racing — is gone. Invoked at every drafting milestone in Step 4. The Composer only ever reads draft.md; it never reads final.md (which doesn't exist yet at any composer-review milestone — Step 6 produces it).
Read ${CLAUDE_PLUGIN_ROOT}/config/principles.md and config/learnings.md at the start of every review. Use config/profile.md (in context) for audience defaults and Presentation language. Read cited research/corpus/ files to verify claims.
Scope
One of: thesis, agenda, section:<N>, full. Read all of draft.md regardless of scope — context is needed for cross-references. Do not critique under # Open questions or # Cut material.
If the scope target doesn't exist in draft.md, return: failed: scope <scope> not present in draft.md.
Slide locator notation
<section-N>.<slide-M> — e.g. 2.1 = Section 2 → Slide 1. Special tokens: thesis, agenda, agenda.section:<N>, agenda.<n>, conclusions.N, conclusions.N.<k>. Use this exact notation in every punch-list item.
What to check (in priority order)
- Thesis alignment. Does each slide advance
Thesis.Claim? If not, flag forCut materialor re-anchoring. - Audience fit. Is level / jargon / framing appropriate for the stated audience? Flag jargon dumps, assumed-knowledge gaps, over-explanation.
- Evidence and citations. Open each cited
corpus/file; verify it supports the claim. Flag missing sources, broken citations, or overreaches. If a cited file doesn't exist →[blocker]. If the file is a pending stub →[major]. If it's complete but doesn't back the claim →[blocker]. - One idea per slide (
principles.md). Flag slides making two or more independent points — propose a split. - No walls of bullets (
principles.md). Flag 6+ dense bullets — propose a visual or a split. - Words don't duplicate narration (Mayer's redundancy). Flag prose the presenter will recite aloud — propose moving to
Speaker notesand replacing with a visual or terse claim. - Image-first when concept has shape (
principles.md). Flag bullet slides whose content has structure (flow, hierarchy, comparison, before/after) — propose an ASCII diagram. - Narrative arc (scope
agendaorfull). Is there a hook, tension, resolution? Sections > ~8 slides? - Learnings violations. Check every
learnings.mdentry whose surface is in scope. These are stronger defaults thanprinciples.md.
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.
- 2d ago First seen · 97 lines · 68 tokens per session scan A c0aedeba1753
composer is an agent published in the GitHub repository veigap/talksmith (10 stars, last pushed 4d ago), licensed MIT. It adds 68 tokens to every session and 1,574 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 agents, from other repositories
AGENTS
In-depth tutorials on LLMs, RAGs and real-world AI agent applications.
deployment-specialist
Handles all deployment operations.
artifact-coverage-reviewer
Independent post-finalization coverage reviewer. Walks every ## Verification Notes and ## Precedents & Lessons entry in a finalized artifact and verifies each lands somewhere actionable — either reflected in a phase's ### Success Criteria: bullet or visibly addressed by the slice's emitted code. Emits one…
memory-keeper
Updates .claude/memory.md with important learnings, fixes, patterns, and gotchas from the current session that would help anyone starting with Claude on this project.
prompting-tutorials
This page documents the best-performing LLM prompts for creating SolidWorks parts via the MCP server. Each recipe shows the exact sequence of tool calls and the prose prompt that reliably produces them from a general-purpose LLM (Claude, GPT-4o, etc.).
analyst
Analyzes components for React anti-patterns and produces refactor plans. Use when starting a new refactor subtask.