Vexa is an open-source meeting transcription system whose bots join Google Meet, Microsoft Teams, and Zoom calls and stream speaker-labeled transcripts through an API. It is for teams that want meeting transcription and downstream knowledge processing either through hosted services or by running the stack themselves. The catalogue entries provide agents and instructions for working with Vexa.
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 instructions/vexa-ai/vexa/agents-mdgit clone --depth 1 https://github.com/Vexa-ai/vexaWrote 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/instructions/vexa-ai/vexa/agents-md)<a href="https://agentmods.dev/instructions/vexa-ai/vexa/agents-md"><img src="https://agentmods.dev/badge/instructions/vexa-ai/vexa/agents-md.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 | $0.03110 | $0.03110 |
| Opus 5 | $0.01555 | $0.01555 |
| Sonnet 5 | $0.00622 | $0.00622 |
| Haiku 4.5 | $0.00311 | $0.00311 |
Grade A, and why
vexa AGENTS.md 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 5d 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 — 193 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — the agent front door (in-repo)
Two constitutions govern everything here — read them on the docs site or in this tree:
- Architecture — the P-book
(
docs/docs/governance/architecture.mdx): what the software must be (P1–P23, enforced by the CI gate suite). Your change must fit these. - Delivery — the D-book
(
docs/docs/governance/delivery.mdx): how change ships — the whole loop (roadmap · PREPARE · TAKE · validation · merge bar · ship bar) in one file.
This file is the actor contract: intake, claim, and how a session behaves in the checkout.
Researching Vexa? Start here
- Found something (bug, gap, docs lie)? File a GitHub issue — every report enters
state: incomingand waits for triage — three days is our target and we are not hitting it (measured 2026-08-20: 26 waiting, median 13 days). Findings are contributions. - Want to contribute? The whole roadmap comes back in one public GraphQL call:
gh api graphql -f query='
query { organization(login: "Vexa-ai") { projectV2(number: 2) {
items(first: 100) { nodes {
content {
... on Issue { number title url state milestone { title } labels(first: 10) { nodes { name } } }
... on DraftIssue { title body }
}
fieldValues(first: 20) { nodes {
... on ProjectV2ItemFieldSingleSelectValue {
field { ... on ProjectV2SingleSelectField { name } } name }
} }
} }
} } }'
Four axes per item: Lane — the business value, named by who feels it (Google Meet · MS Teams
· Zoom · API integrators · Transcription · Recordings · Webhooks & billing · First run ·
Production ops · Feature shelf) — match it to what your contributor actually uses. Human bar
— what validation costs in human terms (Desk check · Operator run · Solo meeting · Small group ·
Crowd 5+ · In-the-loop UX). Setup — the infrastructure floor (None · Lite · Compose ·
k8s/helm). Milestone — the version it gates (currently v0.12.x).
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.
- 5d ago First seen · 193 lines · 3,110 tokens per session scan A f15ef4c7cca2
vexa AGENTS.md is an instructions file published in the GitHub repository Vexa-ai/vexa (2,737 stars, last pushed 2d ago), licensed Apache-2.0. It adds 3,110 tokens to every session, about $0.0155 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 instructions, from other repositories
stenoai CLAUDE.md
Claude Code instructions for stenolabs/stenoai, covering claude.md, architecture, development commands, backend (python) and desktop app (electron).
stenoai AGENTS.md
AGENTS.md instructions for stenolabs/stenoai, covering agents.md, code review rules, bundled binaries with an api, bare exit() in python and platform parity.
backchannel CLAUDE.md
Claude Code instructions for talberthoule/backchannel, covering claude.md, coordination record, current codebase snapshot, build & run and sentrux structural analysis.
backchannel AGENTS.md
AGENTS.md instructions for talberthoule/backchannel, covering agents.md, current codebase snapshot, build & run, sentrux structural analysis and docker compose (primary).
minutes CLAUDE.md
Claude Code instructions for silverstein/minutes, covering claude.md — minutes, project overview, quick start, full build (cli + tauri app) and or manually.
joinly CLAUDE.md
Claude Code instructions for joinly-ai/joinly, covering claude.md, project overview, build & development commands, install all dependencies (run from repo root) and lint (ruff checks all rules by default).