This file is the runbook for the four poller demos, docker/, docker-memory/, archil/, and openshell/. The five webhook-started providers (cloudflare-containers/, cloudflare-worker/, daytona/, modal/, vercel/) share one agent and one environment, declared in webhook-demo/ and created with ant apply --yes . from that directory. Never run ant apply from this directory: it walks into the poller demos and creates their resources too, and the claude-lock.json it leaves here is then found first by every directory below. Each provider's README is its…
Steady4 mentions in 2 weeks
No other project uses it yet
19k55d agoReviewMountCreate a free account to mount
Long-running Node server (Hono). Sentry posts a signed issue / created webhook → src/app.ts verifies it and hands the issue to src/fixer.ts → one Managed Agents session per issue, with the repository mounted as a github_repository resource, the Sentry MCP credential attached through a vault, a dollar budget, the kickoff message in initial_events, and metadata.sentry_short_id as the dedupe key → the agent reads the issue and Seer's analysis through the Sentry MCP server, reproduces, fixes, pushes autofix/<short-id> → it calls…
Steady4 mentions in 2 weeks
No other project uses it yet
19k55d agoReviewMountCreate a free account to mount
Scheduled deployment: cron → Managed Agents session → Sentry MCP search and optional Seer RCA → /mnt/session/outputs/TRIAGE_REPORT.md. Sentry auth is a refreshable mcp_oauth vault credential; the sandbox never holds a token. Provisioning is ./agents/setup.sh (ant 2.0 or later, jq, uv): it records the slugs in ignored sentry-config.json, runs ant apply --yes agents environments vaults, which creates the agent, environment, and vault and records their IDs in claude-lock.json, then runs oauth_setup.py when the vault has no credential for https://mcp.sentry.dev/mcp. Re-runs publish…
Steady4 mentions in 2 weeks
No other project uses it yet
19k55d agoReviewMountCreate a free account to mount
Things that aren't obvious from the docs and tend to cost debugging time. Use the checklist at the bottom when walking someone through the quickstart. Claude Code owns the plugin's credential and does not expose it; the vault owns the scheduled agent's credential and Anthropic refreshes it. Never inspect or copy Claude Code's credential files, and never put a Sentry token in .env, the shell environment, or the agent prompt. An mcp_oauth credential is keyed by mcp_server_url. When a session's…
Steady4 mentions in 2 weeks
34 projects use a version of this
19k55d agoReviewMountCreate a free account to mount
Stateless webhook bridge: Slack app_mention → Managed Agents session (with routing metadata) → session.status_idled webhook → chat.postMessage in-thread. Pull exact shapes from the /claude-api skill's shared/managed-agents-*.md docs. Run the server with bun run dev. Agent provisioning is ant apply --yes agents environments from this directory (ant 2.0 or later; name the directories rather than ., and pass --yes because you have no terminal for its confirmation prompt, or --dry-run to show the plan first): it creates the agent and environment…
Steady4 mentions in 2 weeks
No other project uses it yet
19k55d agoReviewMountCreate a free account to mount
Things that aren't obvious from the docs and tend to cost debugging time. Neither carries the agent's output. Both are signals with IDs. Anthropic's session.status_idled payload is deliberately thin: {type, id}. You follow up with sessions.retrieve(id) (metadata) and sessions.events.list(id) (output). Push the signal, pull the data. On kickoff, set metadata: {slack_channel, slack_thread_ts, slack_team}. When the idle webhook arrives later with only a session ID, retrieve the session, read those keys back, and chat.postMessage to exactly the right thread, with nothing…
Plain text files in a repository that tell a coding agent how the project works: commands to run, conventions to follow and things to avoid. CLAUDE.md, AGENTS.md, cursor rules and skills are the common kinds.
CLAUDE.md or AGENTS.md?
CLAUDE.md is read by Claude Code. AGENTS.md is an open format that Codex, Cursor and other agents read. Many projects keep one and point the other at it.
What is a skill?
A folder with a SKILL.md that describes one capability, such as filling PDFs or reviewing code. The agent loads it only when the task calls for it.
Can I search my own team's files too?
Your agents already can, over MCP, limited to the files you're allowed to read. Searching them from this page is coming.