agentleFS
alphazedealphazede/bearing-lite/skills/onboard-bearing/SKILL.md

onboard-bearing

Skill3 starsChanged 3 months agoApache-2.0View on GitHub

Guided first-run and later configuration of profiles.json. Use for explicit profile setup or configuration change. Do not use to infer settings, store credentials, or write during ordinary Lifecycles.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 2 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If a user lineups.json exists, return MIGRATION_REQUIRED.

    line 27
  2. Very specific:

    Ask for a missing Test Implementer only in tdd.

    line 30
  3. Distinctive:

    Ask settings one at a time within the requested scope, with no preselected value.

    line 37
  4. Very specific:

    Test Implementer and never copies or invents a route from Product Implementer, Test Engineer, Light Implementer, or a retired role.

    line 44
  5. Very specific:

    Test Engineer, Light Implementer, or a retired role.

    line 45
  6. Distinctive:

    Never invent a setting, write the packaged catalog, or keep a lineup fallback.

    line 60

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

onboard-bearing

Inspect profiles.json when present and ask one setting at a time. For first-run or a requested walkthrough, print the welcome and opening question verbatim from references/welcome.md; offer its Lite/Full starting points. Explain each role from its canonical sibling SKILL.md before configuring it, including Integration Engineer's separate planning and execution sessions. Targeted edits skip the tour and ask only missing choices. Recommend without selecting or writing unless explicitly instructed; preserve unaddressed values. Loop: obtain explicit choices → validate → save authorized fields → validate readback → return.

Inputs and match

  • Match: owner-requested first-run setup, configuration change, or migration; a missing profile alone never authorizes writes.
  • Non-match: ordinary Lifecycle dispatch, silent repair of a valid profile.

Algorithm

  1. Resolve <absolute home>/.agents/bearing-lite/profiles.json per ../bearing-lite/references/profiles.md. Never read packaged bytes as user data. Never search, merge, prefer, or fall back to lineups.json.
  2. If a user lineups.json exists, return MIGRATION_REQUIRED. Do not consume it as live configuration. Migration validates the lineup; ask which route to assign to Integration Engineer execution only if that session is enabled. Ask for a missing Test Implementer only in tdd; never copy or invent its route. It then writes a semantically equivalent profiles.json, validates readback, then removes lineups.json. Do not auto-assign Surveyor or any retired role. A retired roles.plan_integrator or roles.planning_and_design route also returns MIGRATION_REQUIRED: final planning/design is the Planning Orchestrator's own procedure. On explicit request, show and remove only that key; preserve every other role, model, fallback, and capability; never copy its route.
  3. Ask settings one at a time within the requested scope, with no preselected value; skip supplied or unchanged choices: named role/session routes and fallbacks, session enablement, development strategy, planning review, assurance cadence, Scribe cadence, concurrency, clean-session, holds, and optional Reverify. When development_strategy.mode is tdd, collect Test Implementer (roles.test_implementer) independently of Product Implementer (roles.implementer). A missing, disabled, or malformed Test Implementer route returns OWNER_DECISION_REQUIRED with an onboard prompt; must ask for the missing Test Implementer and never copies or invents a route from Product Implementer, Test Engineer, Light Implementer, or a retired role. single_implementer stays valid with no test_implementer field. Enabling Coordinator adds value only for a one-wave need (two or more proven-independent packets, shared wave evidence, or aggregate repair ownership); permit an explicit disabled choice. A disabled Coordinator on a true direct packet is not a capability gap. Declining Reverify or its download persists deterministic_verification.reverify.enabled: false for that named profile. Ask whether coverage assist (review.coverage_assist) is enabled and required; an explicit disabled choice is not a gap; its provider, model and credentials stay in its tooling. Use references/verification-tools.md to explain OCR, Reverify, and other optional testing tools and collect explicit choices. For requested delegation configuration, use ../bearing-lite/references/owner-stops.md#coordinator-escalation-and-delegation: explain Coordinator local decisions and two-attempt escalation; store orchestrator-delegation in holds, disabled unless explicitly enabled. Explain that enabled mode needs scoped task authority, exclusions and expiry, preserves protected actions and review/repair limits, and carries frozen decision holds into packets.
  4. Explain Orchestrator transcription as the default when Scribe is omitted/disabled; an enabled Scribe uses one persistent Lifecycle session and resumes from its receipt after a genuine reset. Preserve explicit personal preferences. Every specialist role is optional. Save explicit enabled/disabled choices; configuring a Reviewer does not request review for a task. Write only authorized choices atomically; validate schema readback. Store no credentials or unrequested backups.

Return and recovery

Return READY, MIGRATION_REQUIRED, no_named_profiles, or OWNER_DECISION_REQUIRED with verdict, candidate_ref, changed_paths, tests, findings, and blocker.

Never invent a setting, write the packaged catalog, or keep a lineup fallback.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.