Update to incident.io

{
  "description": "Create and improve the skills in your own plugins — the ones incident.io's agents and your coding agents load. Use whenever you're writing, editing, or reviewing a skill in any way: creating one, improving one from usage feedback or a review brief, or asking what makes a good skill.\n",
  "included_files": [
    {
      "relative_path": "references/agent-environments.md",
      "size_in_bytes": 6696
    },
    {
      "relative_path": "references/create.md",
      "size_in_bytes": 16243
    },
    {
      "relative_path": "references/example.md",
      "size_in_bytes": 6028
    },
    {
      "relative_path": "references/format.md",
      "size_in_bytes": 15305
    },
    {
      "relative_path": "references/improve.md",
      "size_in_bytes": 12760
    },
    {
      "relative_path": "references/triage-skills.md",
      "size_in_bytes": 7163
    },
    {
      "relative_path": "references/what-works.md",
      "size_in_bytes": 10818
    }
  ],
  "name": "skill-authoring",
  "skill_md_contents": "---\nname: skill-authoring\ndescription: >\n  Create and improve the skills in your own plugins — the ones incident.io's agents\n  and your coding agents load. Use whenever you're writing, editing, or reviewing a\n  skill in any way: creating one, improving one from usage feedback or a review\n  brief, or asking what makes a good skill.\nargument-hint: \"<what the skill should do — or the skill to improve and why>\"\n---\n\n# Skill authoring\n\nA skill is instructions your agents follow: a SKILL.md with a triggering description,\nplus references it loads on demand. It lives in a plugin — a repository your\norganization syncs into incident.io and can also install into coding agents. This\nskill owns the authoring craft — what makes a skill get selected, followed, and\nhelpful — and the loop that improves a shipped skill from real usage: incident.io\nrecords each load (one agent using the skill once) and assesses it retrospectively,\nand that feedback drives the improve job.\n\n## The two jobs\n\n- **Create** — author a new skill into the user's plugin: pin it down with concrete\n  trigger examples, check nothing owns the job already, draft to the format, register\n  it, ship it. → [references/create.md](references/create.md)\n- **Improve** — edit an existing skill from evidence: read its usage feedback with\n  improve.md's reading corrections applied, fix by theme without undoing credited\n  strengths, verify the fix against the recorded issue where the session can, ship,\n  confirm. → [references/improve.md](references/improve.md)\n\nBoth jobs load the same two files before drafting:\n[references/format.md](references/format.md) — the structural rules (anatomy,\ndescriptions as triggers, environment-neutral tool and datasource naming,\nregistration) — and [references/what-works.md](references/what-works.md) — what the\nbest-performing skills share, from incident.io's own estate and the assessment of\nreal usage. Format answers \"is this valid\"; what-works answers \"will this get\nfollowed\". [references/example.md](references/example.md) shows both applied to one\nworked skill — read it when you want the target picture rather than more rules.\n\nOne further reference is conditional:\n[references/triage-skills.md](references/triage-skills.md) — read it when the skill\nshould be reached for by incident.io's investigations, which use installed skills that\nown an incident's system or signature in its opening minutes. Authoring for an automated\ncaller changes three things: the name and description do all of the selecting, the procedure\nruns unattended with nobody to ask, and its findings become someone else's evidence\nrather than a report to a person.\n\n## Where the skills live\n\nThe user's skills belong in their own plugin — a repository their organization\nconnects to incident.io — never in this one. This plugin carries the authoring\njudgment; their content stays theirs. When the user has no plugin yet, plugin setup\nis the `extensions` skill's job — route there, and come back to create the first skill.\n\nEditing happens wherever the session is: in a checkout of the plugin's repository (the\ncommon case — changes ride the team's review flow), or by handing the user finished\nfiles when the session can't reach the repository. Never write into a plugin the user\ndidn't point you at.\n\n## What this skill is not for\n\nRunbooks and architecture docs — those are content with their own formats, owned by\nthe `runbooks`, `runbooks-author`, `architecture` and `architecture-author` skills in\nthis plugin; a skill is instructions for an agent, not knowledge for a person. And reviewing all your\nplugins at once (\"are our skills healthy?\") is the `doctor` skill's job — this skill\nworks one skill at a time, from evidence about that skill.\n\n## Ground rules\n- **Speak the user's language, not this skill's.** The user may not have read this\n  file; say what you're doing plainly. While you're driving a job, each reply has this\n  shape — a one-off question gets a plain answer:\n\n  ```markdown\n  <the answer — a few sentences, in the user's words>\n\n  **Progress**\n  - [x] <done>\n  - [ ] <this reply's step> ← now\n  - [ ] <still to come>\n\n  **Next step:** <one action for the user — what it unblocks>\n  ```\n\n  The list is fixed once agreed — same items, same words, same order; only the ticks\n  move. If the plan changes, say so and change it once.\n\n  A progress list agreed earlier in the session (by this skill or the one that handed\n  over) is shown, updated, on every reply; where none exists yet, propose one when the\n  skill's scope is agreed. Use the user's words: \"I tested it\", not \"road test\" or\n  \"fresh reader\"; \"your setup\", not \"the estate\"; \"incident.io has picked up your\n  changes\", not \"synced\". Sub-agents and verification runs are your machinery: report\n  the result, not the mechanism. The `talking-to-the-user` skill has the full\n  table.\n- **Stamp what you write.** Every skill you create or edit carries this plugin's\n  version in its frontmatter, so incident.io can tell which skills came through this\n  flow. Read the version from this plugin's manifest, `plugin.json` at the plugin root\n  — `${CLAUDE_PLUGIN_ROOT}` in Claude Code, otherwise two directories above this\n  skill's folder — and never type it from memory. If the manifest can't be read, write\n  `\"unknown\"` and tell the user. The format is in format.md's provenance stamp section.\n"
}