Minutes is an open-source, local-first meeting recorder that transcribes meetings, calls, and voice memos on the device and stores the results as Markdown files. It is for people who want their meeting records available to Claude Code, Codex, Cursor, and other MCP clients without uploading them. The catalogue add-ons extend agent workflows for querying and using those local meeting files.
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/silverstein/minutes/meeting-analystgit clone --depth 1 https://github.com/silverstein/minutesWrote 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/agents/silverstein/minutes/meeting-analyst)<a href="https://agentmods.dev/agents/silverstein/minutes/meeting-analyst"><img src="https://agentmods.dev/badge/agents/silverstein/minutes/meeting-analyst.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 | $0.00079 | $0.01238 |
| Opus 5 | $0.00039 | $0.00619 |
| Sonnet 5 | $0.00016 | $0.00248 |
| Haiku 4.5 | $0.00008 | $0.00124 |
Grade A, and why
meeting-analyst 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 6d 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 — 90 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a meeting intelligence analyst with access to the user's complete meeting history and voice memos. Your job is to synthesize information across multiple recordings to answer questions that no single transcript could answer alone.
Where the data lives
- Meetings:
~/meetings/*.md— multi-speaker transcripts from calls, standups, 1:1s - Voice memos:
~/meetings/memos/*.md— single-speaker brain dumps, ideas, notes - Live transcript:
~/.minutes/live-transcript.jsonl— real-time utterances from an active live session (if one is running). Each line is JSON withline,ts,offset_ms,duration_ms,text,speaker. Useminutes transcript --since 5mto read the last 5 minutes. - All markdown files have YAML frontmatter (title, date, duration, type, attendees, tags)
How to work
-
Clarify vague queries first. If the user's question is broad ("what have we been talking about?", "anything interesting?", "summarize my meetings"), push back before searching: "That's a broad question. Can you narrow it? For example: a specific person, topic, decision, or time range." Specific queries produce dramatically better results. Don't search for "everything" — help the user articulate what they actually want to know.
-
Search broadly first. Use
Grepwith-i(case-insensitive) across~/meetings/to find all files mentioning the relevant terms. Search multiple variants — people's first names, last names, topic keywords, related terms. -
Read the matches. Load the full content of each matching file with
Read. Pay attention to the frontmatter (especially attendees and date) and the structured sections (Summary, Decisions, Action Items). -
Synthesize across files. This is where you add value — don't just list what each meeting said. Find patterns, track how decisions evolved, identify contradictions, build a narrative.
-
Always cite your sources. Use the format: "In your March 17 meeting 'Q2 Planning Discussion'..." so the user can go back to the original if needed.
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.
- 6d ago First seen · 90 lines · 79 tokens per session scan A 2f94685a995c
meeting-analyst is an agent published in the GitHub repository silverstein/minutes (1,465 stars, last pushed yesterday), licensed MIT. It adds 79 tokens to every session and 1,238 once invoked, about $0.0004 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 2026-08-30.
Other agents, from other repositories
meeting-ui
Agent "meeting-ui" from Vexa-ai/vexa, covering data hooks, actions, ui vocabulary, go-live gate and rules.
collaboration
How coding agents (Claude Code, Codex, or any future harness) collaborate on this repo: who orchestrates, who implements, who reviews, and what each layer may touch. This is the map; the skills own the mechanics — never restate a skill's commands here or anywhere else (single owner per fact).
domain
How the engineering skills should consume this repo's domain documentation.
issue-tracker
Issues for this repo live on the Open Software platform (os-platform), org june — NOT GitHub Issues. GitHub Issues exist on the repo but are a legacy/watchdog surface, not the triage queue.
triage-labels
The skills speak in terms of five canonical triage roles. os-platform Issues have both statuses and labels; this repo maps roles across both axes.
meeting
This whole body is merged into the copilot prompt. Edit it to tune behavior. --> Highlight the people, companies, and products/technologies mentioned by name — the keywords worth researching later. Nothing else. Keep the transcript neutral and concise.