Borrowing it
Nothing to install: this file belongs to LunarWerxs/AgentHydra. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/LunarWerxs/AgentHydra/main/.claude/commands/hydra-status.mdgit clone --depth 1 https://github.com/LunarWerxs/AgentHydraWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/lunarwerxs/agenthydra/hydra-status)<a href="https://agentmods.dev/commands/lunarwerxs/agenthydra/hydra-status"><img src="https://agentmods.dev/badge/commands/lunarwerxs/agenthydra/hydra-status/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/lunarwerxs/agenthydra/hydra-status"><img src="https://agentmods.dev/badge/commands/lunarwerxs/agenthydra/hydra-status.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00023 | $0.00865 |
| Opus 5 | $0.00012 | $0.00432 |
| Sonnet 5 | $0.00005 | $0.00173 |
| Haiku 4.5 | $0.00002 | $0.00086 |
Grade A, and why
hydra-status scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Report the state of the fleet. READ-ONLY: do not archive, surface, deliver, rename, or launch anything, even if something obviously needs it - say what needs doing and stop. Extra focus, if any: $ARGUMENTS
⚠️ ONE PROGRAM, TWO HALVES (owner order, 2026-09-03: the orchestrator moved back into the
AgentHydra repo so nobody has to explain "you have to use both"). The daemon knows what
instances, chats and accounts exist and acts on one when asked. Deciding what should happen
to a chat is the orchestrator's job - the Python toolbox under orchestrator/ in this
repo, driven through the SAME MCP server (orchestrator_menu, orchestrator_run,
orchestrator_loop, orchestrator_switch). So a full status read still takes both halves,
but both are one tool list away. The v2 single-tool answers are gone: prestart,
sweep_loop, deliveries, holds, chat_gate, chat_act, chat_sweep and courier no
longer exist as MCP tools; if you reach for one you are reading an old copy of this file.
Gather - the daemon half:
list_instances- which apps are actually open, and which accounts they are signed into. One or zero open instances means detection is broken, not that the fleet is quiet: say so rather than reporting a quiet fleet.list_usage- every account's remaining quota, each with its own verdict. Name the instance beside every percentage; an unattributed number is worse than none. The weekly (all-models) window is the binding cap except on Pro, where the 5-hour window usually binds first.list_rate_limited_sessions- anything already parked against a wall, and until when.check_update- whether a newer build is available.chat_dossier- only when a specific chat is in question ("what happened to chat X"); it is one query and it beats guessing from a sidebar title.
Gather - the orchestrator half, both read-only (the shell equivalents are python orch.py armed
and python orch.py loop from orchestrator/):
orchestrator_switch {action:"armed"}- IS ANYTHING ALLOWED TO ACT AT ALL. Nothing acts without the tray icon: every acting script asks for its heartbeat first, and the default on any machine is off. Report this FIRST when it is down, because a disarmed fleet looks exactly like a healthy quiet one and every lane below will be reporting refusals rather than results. Do NOT arm it from here - this command is read-only.orchestrator_loop {}- THE DRY LOOP: census, waiting scan, accounts and bands, the sweep's four lanes, naming, reconcile, and the judgment queue, printing what it WOULD do and touching nothing. This is wherestalled,holds,collisions,handoffSoonand the pending deliveries live. STOP AND INVESTIGATE if its census sanity rail fails or the plan reports INCOMPLETE - a read failed, so every lane is a lower bound, not an answer.- the dashboard at http://127.0.0.1:7799 - the same picture in a browser: the accounts strip with the usage bands, every configured rule with its live value, and per-chat dry-run decisions. Cite it rather than re-deriving it by hand.
Then answer in plain English, worst first: what is broken, what is waiting on a person, what is merely queued. One line each. If everything is healthy, say so in one line - do not pad it.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago Changed · +28 lines d72ce328194f
- 10d ago First seen · 24 lines · 23 tokens per session scan A d765288cf1c9
hydra-status is a command published in the GitHub repository LunarWerxs/AgentHydra (43 stars, last pushed yesterday), licensed MIT. It adds 23 tokens to every session and 865 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
ashlr-budget
Set, check, or clear a session spend cap. Guards against runaway tool use in long sessions.
ashlr-eco-mode
Toggle eco mode — aggressive token-saving behaviors that trade some response richness for lower session cost.
ashlr-parallelize
Spawn N parallel sub-agents — one per file — each applying the same task instruction. Merges results into a unified report.
ashlr-tour
60-second guided walkthrough of the ashlr-plugin on the current project — proves the value of the tools on real files.
ashlr-savings
Show estimated tokens and cost saved by the ashlr-plugin this session and lifetime.
ashlr-team-invite
Invite someone to your ashlr team by email. Sends a magic-link invite (admin role required).