your app

events

agentry

prompts

output

App observability for your AI agent.

Send logs, errors, analytics, revenue and deploy events. Your agent investigates issues, fixes bugs and builds operational views and automations.

Install https://agentry.sh and set it up

Built for AI coding agents

Works with any AI coding agent

How it works

Everything is just a prompt away.

Install the agentry MCP and ask your agent to build whatever you need.

Your app

Sends events on every meaningful action — logs, analytics, revenue, deploys.

Agentry

data layer

Stores everything. Queryable via SQL. Webhooks for automations.

Output

Your agent builds you dashboards, fixes bugs, automates reports, what ever you ask for.

claude — ~/myapp

Build me a view to visualise my signup → activation funnel with drop-offs

▾ Worked for 14s · 3 tool calls

● agentry_run_recipe(funnel_3_step)

● agentry_query(events, range:30d)

● Write(src/pages/admin/funnel.astro)

Done. The funnel page is live at /admin/funnel. Biggest leak: email_verified → activated at −59% — users land in the dashboard but don't finish setting up their first workspace.

Signup → activation funnel

Last 30 days · 15,460 visitors

Activation

7.6%

landing_view 15,460 · 100%

↓ −73%

signup_started 4,180 · 27%

↓ −31%

email_verified 2,890 · 19%

↓ −59%

activated 1,180 · 8%

codex — ~/myapp

Why are some users getting "session expired" right after login this week?

▾ Worked for 18s · 5 tool calls

● agentry_list_cases(query: "session expired", range: 7d)

● agentry_get_case(id: 89)

● agentry_query(events, where: "session.duration < 60s")

● agentry_list_deploys(range: 7d)

● Read(src/middleware.ts, src/auth/session.ts)

Found it. Cloudflare is overriding your session-cookie TTL — your /auth/* routes return Cache-Control: max-age=60 from edge cache, but the server sets the session cookie to 30 min. Started 4h ago after deploy 3a9f72 moved cache rules around. 23 events · 8 users · all on the same plan.

Fix: strip Cache-Control on auth routes in src/middleware.ts:42. Want me to push the PR?

yes, push it + add a regression test

▾ Worked for 9s · 3 tool calls

● Edit(src/middleware.ts)

● Write(tests/auth/session-cache.test.ts)

● Bash(gh pr create)

Done. PR #142 opened — strips the conflicting Cache-Control header on /auth/* routes and adds a test pinning the 30-min cookie TTL.

cursor — ~/myapp

Send me a weekly Slack post with DAU, signups, deploys, and MRR

▾ Worked for 9s · 4 tool calls

● agentry_run_recipe(active_users_daily)

● agentry_run_recipe(open_cases_top)

● agentry_query(mrr, weekly)

● cron_create(Mon 9am → format → slack)

Scheduled the weekly digest. Next post lands Monday 9am UTC in #product-health — DAU, MRR, signups, deploys all in one message.

agentry-bot 9:00 AM

#product-health

Week of May 4–11

DAU

1,420 +12%

MRR

$12,400 +8%

Signups

142 +24%

Deploys

14 0 rollbacks

Top funnel leak: email_verified → activated (−59%)

Top open bug: Slack OAuth 502 — 12 events, 4 workspaces

claude — ~/myapp

Ping me on Discord when errors spike above 20/hour with the affected workspaces

▾ Worked for 6s · 2 tool calls

● agentry_create_alert(errors_by_hour, gt:20)

● agentry_register_webhook(discord.com/...)

Alert is armed. Discord webhook signed and verified — you'll get pinged on #oncall the moment errors cross 20/hr.

agentry-bot just now

#oncall

Critical

Error spike — checkout

Rate: 47 / hr (threshold 20)
Started: 4 min ago
Error: PaymentError: Stripe 3DS challenge timeout
src/checkout/stripe.ts:88
Affected: 9 workspaces — acme, linear-clone, finchhq +6

View case Page on-call Suppress 1h

None of these are agentry features. They're things your agent builds on top of agentry's MCP. We store the events; you ship the experience.

Why this exists

The interface changed.

Modern observability tools were designed for humans clicking dashboards. AI agents change the interface.

Agents can:

▸ build customized views
▸ automate operational workflows
▸ generate integrations
▸ analyze trends and suggest improvements
▸ investigate incidents and fix bugs

Agentry is designed from the ground up for agentic software development.

Features

What's in the box.

Built for AI agents as the primary user, not humans clicking dashboards.

ONE EVENT STORE

Logs, errors, analytics, and deploys land in the same project. One key, one dataset, one query surface.

NO SERVER-SIDE MAGIC

agentry stores; the MCP transforms. Stack unmangling, install recipes, every transformation — code sitting in your node_modules. Reviewable in 30 seconds; reproducible offline with the same library.

CASES & SUPPRESSION

Duplicate errors collapse into one case by fingerprint. Your agent records suppression rules — noise teaches itself out.

DEPLOY-AWARE DEBUGGING

Automatically correlate regressions with releases and commits. One tool call answers "what shipped before this broke?".

AGENT-GENERATED OPS

Queries become conversations. Dashboards become customized artifacts. Ask, your agent runs the SQL and writes the page.

WEBHOOKS & AUTOMATIONS

Signed webhooks on any event. Cron jobs your agent writes. Automate fixes, reports, and alerts — your code runs, not ours.

FAQ

Questions you might have.

Compatible with Sentry?

Yes — drop-in. Point your existing Sentry SDK at agentry's /v1/store/{project_id}/. The wire format is Sentry's literal event schema (event_id, exception.values, stacktrace.frames). Auth accepts X-Sentry-Auth and sentry_key.

Compatible with PostHog?

Yes — and PostHog is the analytics backend under the hood. Each agentry user gets a provisioned PostHog project. PostHog-shaped clients drop in at /v1/track/. Your analytics keep working; you just get an agent-first interface on top.

Do I need to use Claude Code?

No. Agentry speaks MCP, so it works with any MCP client — Cursor, Windsurf, Cline, Codex, or your own. We optimize for Claude Code because that's what we use, but nothing's tied to it.

What languages does this support?

Anything that can POST JSON. There is no SDK to install — your agent generates a 25-line fetch helper at install time, tuned to your stack. Reviewable in 30 seconds, no vendor dependency to vet, no upgrade cycle.

Does the server transform my data?

No. agentry's HTTP API is the data plane — storage, retrieval, queries. It never translates, normalizes, or rewrites anything. Transformations (minified stack unmangling, fingerprinting, formatting) run locally in your MCP process where the code is on npm and sitting in your node_modules. If a translation looks wrong, you can read the exact code that produced it and reproduce it offline with the same library. No server-side magic.

What about humans who want a dashboard?

Your agent writes the dashboard as a real page in your repo. You can commit it, edit it, and ship it. No SaaS UI to learn — and no permanent dashboard config rotting over time.

Do you train AI on my data?

No. We never run LLMs server-side. The agent that reads your data is your Claude Code (or other MCP client) running on your machine. Your data stays in your agentry project.

agentry isn't trying to replace Sentry's source-map polish or PostHog's cohort-analysis depth. We sit on top of both worlds with a single ingest, an MCP-shaped output, and a workflow that ends in a PR instead of a notification.

Install