agentburn 0.5.1

Where does your AI agent burn money? Local token/cost profiler for always-on agents (Hermes Agent first). Zero dependencies.

agentburn

Where does your AI agent burn money — while you sleep?

Always-on agents bill you around the clock. Hermes Agent users wake up to $47 overnight bills from recursive subagent runs; one user measured that 73% of every API call is fixed overhead (tool definitions + system prompt, resent every time); chained delegation means "step 3 costs 4× step 1 — no alert, just a bill." Built-in /usage shows totals. Nothing shows where it burns.

agentburn is a local profiler for your agent's own accounting data — universal across agents: Hermes Agent, OpenClaw and Claude Code today, one normalized model underneath. One command, zero dependencies, nothing leaves your machine:

uvx agentburn          # or: pipx run agentburn / pip install agentburn

🔥 agentburn — hermes · last 30d

   TL;DR: ≈ ~$431/mo pace; 79% of it is `cron`.
   First fix: 79% of spend happens at night — route night work to a cheaper model

   ~$45.50 total · 1.75M tokens · 7 sessions · 123 API calls
   ≈ ~$431.24/month at the current pace

   WHERE IT BURNS (by source)
   cron                 ██████████████····  79%     ~$36.00    1.24M  2 sess
   cli                  ██················   9%      ~$4.00     185K  1 sess
   gateway:telegram     █·················   7%      ~$3.00     210K  1 sess
   subagent             █·················   5%      ~$2.50     113K  2 sess

   🌙 WHILE YOU SLEPT (00:00–08:00): ~$36.00 (79% of spend) · 2 sessions
      mostly: cron

   FIXED OVERHEAD (avg input tokens per API call)
   gateway:telegram       20,000 ← heavy
   cron                   15,000 ← heavy
      input composition (sampled from 3 request dumps): system 30% · tools 58% · history 12%

   💡 DO THIS
   1. 79% of spend happens at night — that's ≈$341/mo while you sleep. Route night work to a cheaper model.
   2. Scheduled (cron) sessions run on anthropic/claude-opus-x — maintenance rarely needs a frontier model.
   3. 20,000 input tokens per call on telegram: trim per-platform toolsets, prune unused skills.

What it answers

• Where it burns — by source: cron / subagent / gateway:telegram|discord|whatsapp / cli. Always-on ≠ free: scheduled jobs and gateways spend without you.
• 🌙 While you slept — the overnight bill, isolated and named (configurable window: --night 23-7).
• Fixed overhead — average input tokens per API call per source. The "73% overhead" pattern is visible in one glance; with request dumps enabled, you get the sampled composition (system prompt vs tool definitions vs history).
• Subagent rollups — delegation cost chained back to the session that spawned it. Recursion compounds; here is the receipt.
• Top tools — which tool results weigh most in your context.
• What to do — up to 4 conservative, named recommendations with monthly estimates.

Why trust these numbers

Most token trackers quietly disagree with each other (2–91× in public issue threads). agentburn takes the opposite stance:

• Numbers come from the agent's own accounting (~/.hermes/state.db: per-session token counters and cost fields). No scraping, no proxies, no guessing.
• Provider-billed costs are shown as-is; Hermes estimates are marked with ~. Mixed data is labeled mixed.
• Sessions with messages but zero recorded tokens (known Hermes accounting gaps, e.g. #12023) are detected and reported: totals are then explicitly a lower bound — and fixing the accounting becomes recommendation #1.
• Input composition from request dumps is char-proportional and labeled sampled estimate, not truth.

Privacy

Everything runs locally and reads your database read-only. No network calls. No telemetry. The report is yours.

Usage

agentburn                        # every agent on this machine, last 30 days
agentburn --agent openclaw       # just one
agentburn --days 7
agentburn --agent hermes --db /path/to/state.db
agentburn why                    # behavioral forensics: loops, retry storms, idle heartbeats
agentburn --night 23-7           # custom overnight window (local time)
agentburn --budget-month 50 --fail-over   # sentinel for cron/CI
agentburn --json                 # machine-readable, pipe it anywhere
agentburn --no-color

Mechanics

📤 Share your burn (--share). An anonymized card — categories, models and totals only; session titles, paths and content are excluded by construction. Safe to paste into a post; --svg card.svg renders the same card as an image:

🔥 my hermes agent · last 30d
~$45.50 → ~$430/mo pace · 1.75M tokens
where it burns: cron 79% · cli 9% · telegram 7% · subagent 5%
🌙 while I slept (00–08): ~$36.00 — 79% of everything
⚙️ telegram re-sends 20,000 tokens with EVERY call — 2.5× the community norm (≈8k)
— agentburn · local & private

--svg card.svg renders it as an image:

📏 Calibration against public benchmarks. "Is 15k input tokens per call normal?" The report compares your fixed overhead with community-measured references embedded as dated constants (e.g. the Phala always-on-agent benchmark, 2026-03: ≈8k/call baseline). No network — sources are cited inline.

📐 Optimize → prove it (--save-baseline / --compare). Snapshot your pace, change the config (cheaper cron model, trimmed toolsets), then agentburn --compare shows the delta in $/month — pace-normalized, so a 7-day baseline compares honestly with a 30-day window. Every recommendation becomes a testable promise.

🔬 agentburn why — behavioral forensics. report says where it burns; why says why, from the agent's own recorded actions and thoughts:

🔬 agentburn why — openclaw

   RE-READ LOOPS          5× browser(https://news.site/page) — every repeat re-paid in full
   RETRY STORMS           Bash: 3 errors / 6 calls — paying full price for every error
   IDLE HEARTBEATS        4 of 9 heartbeat runs did NOTHING — $2.40 of pure idle burn
   BURNED ON FAILURES     2 failed runs → ~$3.90 (timeout, killed)
   THINKS MORE THAN IT WORKS   62% thinking · 84K tokens · "rename files task"

   💡 WHAT TO CHANGE
   1. `/proj/big.md` was fetched 4× in one session ≈32K tokens re-paid — cache it…

Observations with numbers, not verdicts; only tool names, truncated argument keys and counters — message content never leaves the machine (and never enters the report).

🩺 agentburn doctor. Trackers disagree because the agent's own accounting has gaps. doctor names the broken combinations (provider × model × source) for zero-usage and unpriced sessions, and generates a ready-to-paste upstream bug report — counters only, no message content.

🚨 Sentinel mode — a budget guard for server agents. Your agent runs 24/7 on a VPS; this watches it:

# alert when overnight burn exceeds $5/month pace (exit code 1 → any alerting hooks in)
agentburn --agent openclaw --budget-night 5 --fail-over --no-color \
  || notify-send "🚨 agent is burning money at night"

Drop it in cron next to the agent itself — the one-off check becomes a standing guard.

Supported agents

One normalized model, one adapter per agent. Run agentburn and every agent found on the machine gets its own report.

Agent	Status	Data source	Notes
Hermes Agent	✅	~/.hermes/state.db (+ optional request dumps)	costs from the agent's own accounting
OpenClaw	✅	~/.openclaw/agents/*/sessions/sessions.json	heartbeat is its own category — the famous one; cron / gateways / subagents split out
Claude Code	✅	~/.claude/projects/**.jsonl	tokens only, by design: CC doesn't record costs locally and subscription usage has no honest per-token price — we don't invent one

Adapters are ~150 lines over a shared model. Codex CLI / opencode are natural next targets — PRs welcome.

Related

token-history — the macro view: daily archive of which agents the world uses (OpenRouter rankings). agentburn is the micro view: where yours burns.

License

MIT