# amplifthq/opentag

Mention any ACP coding agent from Slack, GitHub, GitLab, Linear, or Lark. OpenTag runs Claude Code, Codex, Cursor and more on your own machine, then replies in-thread with verified, evidence-backed results.

Repository: https://github.com/amplifthq/opentag
Canonical: https://ross.abutalabs.com/products/opentag
Language: TypeScript
License: MIT
License Family: permissive
Topics: ai-agents, codex, github, slack, acp, agent-client-protocol, claude-code, coding-agents, developer-tools, local-first
Last push: 2026-08-24T01:36:06+00:00

## Health v2 (maintenance only)
Score: 80/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 98, longevity 5
- inputs: {"age_days": 70, "days_push": 10, "days_rel": 16, "gap_med": 4, "n_releases_24m": 8}
- flags: young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1369, forks 78 (observed 2026-08-28T04:04:31.543640+00:00)

## What it is
OpenTag is a local-first gateway that lets you mention coding agents like Claude Code, Codex, and Cursor from Slack, GitHub, GitLab, Linear, or Lark, running them on your own machine via the Agent Client Protocol. It replies in-thread with evidence-backed results and enforces completion gates such as merged pull requests before marking work done.

## Use cases
- trigger claude code from slack
- run codex agent from a github comment
- delegate coding tasks to ai agents from linear
- verify ai agent work with evidence before merging
- run coding agents locally without sending code to the cloud
- orchestrate multiple coding agents through one protocol

## When to choose
- your team collaborates in Slack, GitHub, GitLab, Linear, or Lark and wants to dispatch coding agents from threads
- you need local-first execution so code and credentials never leave your machine
- you want auditable, evidence-gated completion instead of trusting agent self-reports
- you want vendor-neutral agent dispatch across multiple ACP-compatible executors

## When to avoid
- you need a fully hosted SaaS with no local component
- you don't use any of the supported chat or code platforms
- you only need a single agent integrated directly without a gateway
- your platform of choice (Teams, Discord, Telegram) is still in preview

## Facets
- artifact type: service
- maturity: active
- function: agent-framework, workflow-automation, developer-tools, webhook, chatbot
- domain: developer-tools, large-language-models
- platform: cli, self-hosted, cross-platform
- tags: agent-client-protocol, acp, claude-code, codex, cursor, slack-integration, github-integration, local-first, coding-agents, audit-ledger, ai-agents, automation, nodejs

## Member repositories
- amplifthq/opentag (main) score 80

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:31.543640+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:40:57.795665+00:00, confidence not recorded.
  - readme: https://github.com/amplifthq/opentag (fetched 2026-08-28T04:04:31.543640+00:00, sha f07cd528934a)
- Data as of 2026-08-30T08:39:29.467469+00:00.