intentos 0.5.0

Open-source flight recorder for AI agents — see exactly what your agent did, why it failed, and what it cost

Intent OS

Your AI agent ran for 20 minutes. It says "done."

Can you explain what it did?

---

pip install intentos

# What just happened?
intent-os doctor

# See every step your agent took
intent-os inspect latest

# Track what it cost
intent-os cost

You get this:

[14:02:01] > START
[14:02:09] > MODEL CALL  claude-sonnet-4  (2,451 tokens)
[14:02:14] > TOOL        filesystem.write
[14:02:27] !! FAILED     test_jwt_verify failed

Goal:        refactor-auth-module
Agent:       claude-code
Duration:    14.3s
Cost:        $0.08
Tokens:      4,891

No more guessing. No more "I think it called the API three times." You see exactly what happened — every model call, every tool use, every failure, every dollar.

---

You're not alone

You've been here:

• "What did it actually do?" — Claude Code says "task complete." The file changed. But you didn't see it happen. You don't know if it wrote 3 lines or deleted a function.
• "Why did it fail?" — Agent runs for 30 minutes. Fails. No stack trace. No log. Just "error."
• "Where is the money going?" — API bill shows $47 this month. Which agent? Which model? Which task?

Intent OS is the flight recorder for AI agents. It intercepts every API call your agent makes and turns it into a structured, searchable execution trace. Your data stays on your machine. No cloud. No account. Just pip install.

---

Works with your agent in 30 seconds

# Start the recorder
intent-os proxy start

# Point your agent at it
export OPENAI_BASE_URL=http://localhost:8377
export ANTHROPIC_BASE_URL=http://localhost:8377

# Use your agent normally — every call is recorded
claude "refactor this module"

# See what happened
intent-os doctor
intent-os inspect latest

Works with Claude Code, Cursor, GitHub Copilot, or any agent that speaks OpenAI or Anthropic APIs. Zero changes to your agent. Just one environment variable.

---

What you get

Command	What it tells you
intent-os doctor	One-command health check: what your agent did, what went wrong, how to fix it
intent-os inspect latest	Full execution timeline: every model call, tool use, cost, and duration
intent-os cost	Spending breakdown: by agent, by model, daily trends
intent-os proxy start	Start recording — intercepts Claude Code, Cursor, any agent
intent-os proxy doctor	Check proxy health: running status, traffic stats, agent detection
intent-os agent create --name "My Agent"	Register agent identity for tracking across sessions
intent-os scan	Security scan: detect dangerous tool calls and sensitive data in traces
intent-os audit report --format html	Compliance report for teams: full audit trail with HTML/CSV export
intent-os event prune --older-than 90	Data lifecycle: clean up old traces, keep your disk under control

---

For teams

As your team grows:

• Cost tracking — intent-os cost --by agent — who's spending what, on which model
• Security policies — define what agents can and can't do: intent-os security policy apply
• Compliance audit — full execution records: intent-os audit report --format html
• Agent identity — every agent gets an ID, every execution links back to its owner

---

Why local-first?

Instead of...	Intent OS is...
Cloud-only tracing (LangSmith, LangFuse)	Local-first. Your data never leaves your machine.
Siloed per-platform logs	Universal. Works with any OpenAI/Anthropic agent.
Just logging	Structured traces. One execution → many API calls → one timeline.
Postgres + Redis + S3	One SQLite file. No infrastructure needed.

No API key to sign up. No dashboard to log into. Your agent's execution data is yours — it lives in ~/.intent-os/events.db.

---

Architecture

AI Agent → Intent OS Proxy → LLM (OpenAI / Anthropic / Ollama)
                │
                ├── Flight Recorder (observe / debug / cost)
                ├── Security Guard (scan / policy / audit)
                └── Event Store (SQLite — local, append-only, queryable)

Intent OS is building the execution layer for AI agents. Today it's a flight recorder. Tomorrow it will make agents portable, governable, and composable across any runtime.

---

Tested

731 tests, 8 skipped, 0 failures — CI across Python 3.10, 3.11, and 3.12.

---

License

AGPLv3 + Commercial Option. See LICENSE.

Open-source use is free under AGPLv3. Commercial use requires a commercial license.

---

Built by one person, for every developer who's asked "what did my agent just do?"