Check that this machine's agent team can work and say what is wrong in plain lines: tools installed, core skills installed, skill commands runnable, hooks configured, Reach seat live, a heartbeat running, both vendors' quota readable, temp space free. Checks only, never repairs. Use on a new machine, when agents stop reaching each other or a long run fails to start, or when the owner asks whether everything is healthy.
The English summary is being prepared. Showing the original description.
Run score Not tested
See how to installReview the instructions the agent will receive before you install.
sno medic run looks at the agent team on this machine and prints one line per check. It reads and
reports; it never installs, repairs, starts, stops or spends anything.
OK <check>: <what was found>
WARN <check>: <what is wrong> -> <the command that fixes it>
FAIL <check>: <what is wrong> -> <the command that fixes it>
MEDIC ok=N warn=N fail=N
FAIL means the team cannot work at all (a required program is missing, no agent command line is
installed, Reach is broken). WARN means it works worse than it should. Exit status is 0 when there is
no FAIL, 1 otherwise, 2 for a wrong command line.
Prerequisites: bash 4+, GNU coreutils and jq. sno medic with no arguments, or --help, prints the usage.
sno medic run. It takes about ten seconds because it asks sno doctor and reads both quotas.WARN and FAIL line with its fix.
Do not paste the whole output when everything is OK; one sentence says so.sno medic run again and read the same line: it must now say OK.| Check | Looks at | Common fix |
|---|---|---|
tools | sno, jq, tmux, git are on PATH, and sno heartbeat and sno subscription-quota-check run | install the named program |
agent-cli | claude or codex is on PATH | install one of them |
skills | the six core skills (reach, heartbeat, join-talk, handoff, deliver, charter) exist where each installed agent reads skills | reinstall the skills |
commands | sno deliver-proof, sno handoff-checkpoint, sno rotate-agent-resume and sno rem-reflect answer --help | reinstall the skills |
hooks | sno doctor reports hooks configured (and trusted, for codex) for each installed agent | reinstall the skills |
reach | sno doctor reports Reach as ok | the fix sno doctor names |
skill-files | every installed skill file passes sno doctor | update the skills |
seat | this agent's own seat (SNO_REACH_ADDR) is live, or at least one seat is live | run join-talk in this window |
heartbeat | at least one heartbeat is running on this machine | arm one when a long job starts |
quota | each installed agent's subscription quota can be read and is not wait | log in, or wait for the reset |
temp-space | free space in $TMPDIR (or /tmp) is at least 500 MB; change it with --min-free-mb N | free space or point TMPDIR elsewhere |
A check that cannot run because a program is missing prints WARN <check>: not checked, <program> is missing; fix the tools line first.
sno medic does not prove that an agent can answer a prompt or that a seat delivers a message: for a live
receiver test before a vendor switch use rotate-agent-preflight (the rotate-agent skill), and for a
single unit's behaviour use that unit's own self-test.
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Answer how long agent work will take, how much longer active work needs, or what clock time applies. Estimate full AI-agent wall time, measure active progress first, and get owner-facing times from `sno report-time`. Never estimate human-developer time.
The English summary is being prepared. Showing the original description.
Prove that a reported run caused its intended real effect. Walk the executor path from entry to effect, capture one observable per hop, and report a result a no-op could not produce.
The English summary is being prepared. Showing the original description.
Give the owner one page when they come back from sleep, a meeting or a long run: what got done, what is stuck, what waits for their answer, and what it cost in quota. Use when the owner asks what happened while they were away, wants a status page, or when you finish a long stretch of work and must report it.
The English summary is being prepared. Showing the original description.
Report how many problems the two brains found: mutual review (the other model checks the work, from the peer-review records) and self-check (an agent finds its own error in its own conversation, judged by that agent's own model), side by side. Use when the owner asks whether the second agent or the reviews are worth it, or wants the week's review results.
The English summary is being prepared. Showing the original description.
Write, revise or review a charter: the one document that tells an executor what to deliver, what is out of bounds, what the owner already decided, and how each success check will be proven. Use when work needs a written brief before someone else does it, or when asked to write, change or check a charter.
The English summary is being prepared. Showing the original description.
Route production coding tasks to the user-level Shell, TypeScript, Python, or Rust Codex workflow. Use for "codex-coder", "Codex write", "generate with Codex", language migrations, or mixed Shell/TypeScript/Python/Rust implementation when Codex should choose the correct language specialist automatically.
The English summary is being prepared. Showing the original description.