io.github.rayson-tech/nomarmy

What it mainly works on, as our AI model read its tool descriptions: files and code on the machine it runs on.

9 of 17 tools are described as changing something (creating, editing, deleting, sending or running).

• local_workerruns code or commandsdescribed as changing data

  “Run one isolated local worker and wait for it.”

  25 inputs, 1 required

  • taskstringrequired“implement: the OBJECTIVE the worker must achieve, not the edit it should make.”
  • criteriaarray“implement: contract IDs from acceptance/*.yml served by this job, read at base_ref when set (else the working tree).”
  • acceptancearray“implement: acceptance criteria the worker must satisfy.”
  • verificationstring“Verification profile NAME (e.g.”
  • refactorboolean“implement: declares a behavior-preserving change (moving or restructuring code).”
  • verify_regressionboolean“implement only: after the diff passes `verification` and touches production files, temporarily revert just those production files, re-run the SAME verification…”
  • modestring“verify: run a required verification profile with no worker and no model tokens; base_ref selects the branch or commit (default current HEAD), task is a short r…”
  • base_refstring
  • timeout_secondsinteger“Default 600; 1200 for a review scout (one with `reviews`, or an army role in the review phase), since a real security review read for the full 10 minutes and w…”
  • reasoningstring“Thinking level passed to the worker model: off, minimal, low, medium, high, xhigh, adaptive, max or ultra (the higher ones only where the vendor offers them, e…”
  • agentstring“Run on this agent from the operator's agents.yml, by name (e.g.”
  • modelstring“The model to run on the job's agent (an api or subscription agent), e.g.”
  • run_idstring“The /feature run this job belongs to (from run_start).”
  • reportstring“How much the worker may report back, capped by its agent's tier: brief (today's local-sized report), standard (the default), full (the frontier ceiling: about …”
  • stakesstring“implement: how much a mistake would cost, separate from how hard the work is.”
  • reviewsstring“scout: the job id this scout independently reviews, so the review is recorded against that job (nomarmy stats shows high-stakes jobs with and without one).”
  • commit_subjectstring“implement: the subject line of the commit nomArmy makes on the worker branch, e.g.”
  • army_rolestring“Dispatch by army role (e.g.”
  • confirm_over_limitboolean“Override a reached usage limit: the General must ask the operator before resubmitting with confirm_over_limit: true, or send the job to another agent.”
  • on_behalf_ofstring“Required when the job's agent is a subscription: must exactly match that agent's owner in agents.yml, or nomArmy refuses the job.”
  • evidencestring“implement only: facts YOU already resolved (e.g.”
  • reference_commitsarray“Earlier changes this job's worker reads as patch files under .openclaw/reference/ (git history isn't available in the sandbox): a job id (that job's change aga…”
  • continue_fromstring“implement only: the job id of a retained, uncommitted implement job (partial, blocked, or failed verification) whose unfinished work this job should finish.”
  • detachboolean“local_worker_start only: run the job in its own process instead of this session's MCP server, so it keeps running (and records its result) after this session e…”
  • worker_idstring
• local_worker_startruns code or commandsdescribed as changing data

  “Start one worker or scout in the background and return immediately with a job_id.”

  25 inputs, 1 required

  • taskstringrequired“implement: the OBJECTIVE the worker must achieve, not the edit it should make.”
  • criteriaarray“implement: contract IDs from acceptance/*.yml served by this job, read at base_ref when set (else the working tree).”
  • acceptancearray“implement: acceptance criteria the worker must satisfy.”
  • verificationstring“Verification profile NAME (e.g.”
  • refactorboolean“implement: declares a behavior-preserving change (moving or restructuring code).”
  • verify_regressionboolean“implement only: after the diff passes `verification` and touches production files, temporarily revert just those production files, re-run the SAME verification…”
  • modestring“verify: run a required verification profile with no worker and no model tokens; base_ref selects the branch or commit (default current HEAD), task is a short r…”
  • base_refstring
  • timeout_secondsinteger“Default 600; 1200 for a review scout (one with `reviews`, or an army role in the review phase), since a real security review read for the full 10 minutes and w…”
  • reasoningstring“Thinking level passed to the worker model: off, minimal, low, medium, high, xhigh, adaptive, max or ultra (the higher ones only where the vendor offers them, e…”
  • agentstring“Run on this agent from the operator's agents.yml, by name (e.g.”
  • modelstring“The model to run on the job's agent (an api or subscription agent), e.g.”
  • run_idstring“The /feature run this job belongs to (from run_start).”
  • reportstring“How much the worker may report back, capped by its agent's tier: brief (today's local-sized report), standard (the default), full (the frontier ceiling: about …”
  • stakesstring“implement: how much a mistake would cost, separate from how hard the work is.”
  • reviewsstring“scout: the job id this scout independently reviews, so the review is recorded against that job (nomarmy stats shows high-stakes jobs with and without one).”
  • commit_subjectstring“implement: the subject line of the commit nomArmy makes on the worker branch, e.g.”
  • army_rolestring“Dispatch by army role (e.g.”
  • confirm_over_limitboolean“Override a reached usage limit: the General must ask the operator before resubmitting with confirm_over_limit: true, or send the job to another agent.”
  • on_behalf_ofstring“Required when the job's agent is a subscription: must exactly match that agent's owner in agents.yml, or nomArmy refuses the job.”
  • evidencestring“implement only: facts YOU already resolved (e.g.”
  • reference_commitsarray“Earlier changes this job's worker reads as patch files under .openclaw/reference/ (git history isn't available in the sandbox): a job id (that job's change aga…”
  • continue_fromstring“implement only: the job id of a retained, uncommitted implement job (partial, blocked, or failed verification) whose unfinished work this job should finish.”
  • detachboolean“local_worker_start only: run the job in its own process instead of this session's MCP server, so it keeps running (and records its result) after this session e…”
  • worker_idstring
• local_worker_statusreads or searches

  “Status of one job, started by this session or any other (a detached job included): phase (starting, worktree, worker, verification, commit, record, finished), elapsed time against its timeout, and th…”

  4 inputs, 1 required

  • job_idstringrequired
  • wait_secondsinteger
  • fullboolean
  • reportboolean“Just the report and nomArmy's verdict on it, without the rest of the record.”
• statson 28 other serversreads or searches

  “What nomArmy's own job records show for this repository (or all repositories): jobs by mode, role and model; code committed; worker and job time; tokens and API spend; claim vs evidence (how often a …”

  8 inputs

  • sincestring“Only jobs started since this: a date (2026-09-25) or an age (7d, 24h).”
  • untilstring“Only jobs started before this date.”
  • all_reposboolean“Every repository this machine's nomArmy has run jobs for, not just this one.”
  • repostring“Another repository, by path or folder name ("senti" matches rayson-senti if unambiguous).”
  • rolestring“Only jobs dispatched as this army role (e.g.”
  • modelstring“Only jobs that ran on this model (e.g.”
  • formatstring“text (default) is the report; json is the raw numbers.”
  • detailsboolean“The full report (volume, reviewers, flags, what didn't finish).”
• local_worker_stopruns code or commandsdescribed as changing data

  “Stop a running job's worker, for example one burning a frontier model's usage on the wrong track.”

  2 inputs, 1 required

  • job_idstringrequired
  • reasonstring“Why it's being stopped; recorded in the job's issues.”
• local_worker_capacityreads or searches

  “What this host can take right now: context per nom and the brief/report budgets derived from it, memory pressure and whether another job would be admitted, and the jobs currently running.”

  Takes no inputs.

• run_starton 1 other serverruns code or commandsdescribed as changing data

  “Start a /feature run (or reattach to one with `resume`): one feature, end to end, with limits.”

  5 inputs

  • namestring“A short name for the feature (required unless resuming).”
  • resumestring“Reattach this session to an existing, still-running run (e.g.”
  • max_jobsinteger
  • max_api_usdnumber
  • max_hoursnumber
• run_statuson 3 other serversreads or searches

  “A /feature run's limits, what it has used (jobs, api spend, hours), per-agent jobs/spend/tokens, warnings (80% of a limit, paused agents), and its log path.”

  1 input, 1 required

  • run_idstringrequired
• run_finishruns code or commandsdescribed as changing data

  “Close a /feature run as complete or stopped, with a one-paragraph summary.”

  3 inputs, 3 required

  • run_idstringrequired
  • statusstringrequired
  • summarystringrequired
• armyreads or searches

  “Who you, the General, are and who you call for what in this repository: your fixed charter and the agent you're defined as, the army's workflow, then each role's description, phase (build, review, ac…”

  Takes no inputs.

• local_worker_configreads or searches

  “What this checkout's .nomarmy.yml defines -- the one file every job's sandbox image and every verification uses, whichever branch the job starts from (never the job worktree's own copy, which a worke…”

  Takes no inputs.

• local_workersruns code or commandsdescribed as changing data

  “Run independent jobs (implement or scout) with bounded parallelism and wait for all of them.”

  4 inputs, 1 required

  • jobsarrayrequired
  • max_parallelinteger“A cap on how many of this batch run at once.”
  • auto_unionboolean“After all jobs finish, mechanically merge (git merge --no-ff) implement jobs that reached a valid outcome and touched non-overlapping files into ONE new integr…”
  • union_verificationstring“Verification profile NAME to run once against the union branch after merging (same semantics as each job's own `verification` field).”

Each quoted sentence is the first sentence of the publisher’s own description. The kind of tool and the “changes data” mark are our AI model’s reading of that description, not something we tested: two differently worded runs had to agree, otherwise no label is shown, and nobody has checked them by hand. A tool without the “changes data” mark is described as only reading, searching or calculating, or the two runs did not agree. No tool was called. Read the publisher’s full description before relying on it. Inputs are the parameters in the tool's own input schema: name, declared type, whether it is required, and the first sentence of its description. Read 2026-10-23.

The registry verifies a package once, when it is published, and does not re-check it. A difference below is about the listing, not a judgement of the publisher.