agentleFS
ByronScottJonesByronScottJones/Agents/skills/twg-operational-health/SKILL.md

twg-operational-health

Skill7 starsChanged 2 months agoMITView on GitHub

Use with the root `twg` skill for on-call handoffs, incident response and investigation, post-incident reviews, reliability reviews, Assets refresh, capacity views, meeting summaries, and operational risk readouts.

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. About 0 are common, 8 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

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

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

    Use together with the root twg skill.

    line 11
  2. Very specific:

    Run twg <command>. On shell command not found, use $HOME/.local/bin/twg (macOS/Linux) / $env:LOCALAPPDATA\Programs\twg\bin\twg.exe …

    line 16
  3. Very specific:

    Run without --site; TWG inherits the user's pinned Jira/JSM site.

    line 38
  4. Distinctive:

    Keep operational state distinct from adjacent PR, pipeline, document, or chat activity.

    line 46
  5. Distinctive:

    After one repeated backend, auth, or schema error, report the gap and use remaining evidence instead of aliases or broad inventories.

    line 54
  6. Very specific:

    Read build health from bitbucket pipeline query, not per-pipeline get.

    line 56
  7. Very specific:

    Use for active, newly mitigated, or pre-PIR incidents.

    line 70
  8. Very specific:

    If fields are sparse, probe four golden-signal families with bounded follow-ups.

    line 73
  9. Distinctive:

    Use after mitigation/recovery when drafting or evaluating a PIR.

    line 80
  10. Distinctive:

    Do not fetch every transcript/page body or treat every incident mention as a live risk.

    line 114
  11. Distinctive:

    Do not wait for chat/comments to surface directional pre-PIR hypotheses.

    line 116
  12. Distinctive:

    Do not call mitigation root cause without a mechanism or treat workflow panels, bot comments, or opaque fields as RCA narrative.

    line 117
  13. Distinctive:

    Do not use keyword matches alone as ownership.

    line 119
  14. Distinctive:

    Do not join Assets by display-name guesses or recommend staffing from activity counts alone.

    line 121

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

twg-operational-health

Use together with the root twg skill. Exact command grammar comes from live twg help or twg help describe <path>.

CLI launcher fallback

Run twg <command>. On shell command not found, use $HOME/.local/bin/twg (macOS/Linux) / $env:LOCALAPPDATA\Programs\twg\bin\twg.exe (PowerShell), then tell user to add that directory to PATH. Do not treat auth or command errors as PATH failures.

Use When

  • "I'm taking over on-call"
  • "Reliability, incident, SEV, or post-incident review readout"
  • "Investigate an active incident and find mitigation"
  • "Analyze root cause or draft postmortem/PIR learning/action items"
  • "Open risks, blockers, overloaded people, operational health"

First Move

Resolve scope, window, anchors, owner/escalation, status, recency, and follow-ups before joining surfaces. For incident, PIR, reliability, or on-call requests, start with native JSM incident, post-incident-review, or Jira workitem records when available. Use Rovo or document search only to disambiguate an anchor or fill a bounded gap, never as the incident inventory. Current escalation, owner, and readiness claims need current/open evidence; closed records provide recurrence context, not present-state proof. Run without --site; TWG inherits the user's pinned Jira/JSM site. Only add a site override when the user explicitly requests another tenant. Never assume a vendor-internal incident site.

Evidence Policy

  • Rank by impact, urgency, owner clarity, recurrence risk, and actionability; separate live risks from historical mentions.
  • Keep operational state distinct from adjacent PR, pipeline, document, or chat activity; missing or inaccessible state is an evidence gap.
  • Cluster incident, PIR, follow-up, runbook, owner, asset, or meeting evidence by service, theme, owner, and recency. Hydrate highest-risk clusters and stop when theme, owner signal, and confidence are clear.
  • Separate working theory, confirmed problem, mitigation, and root cause; require a causal mechanism for root cause and label claims confirmed, supported, candidate, or missing evidence.
  • After one repeated backend, auth, or schema error, report the gap and use remaining evidence instead of aliases or broad inventories.
  • Read build health from bitbucket pipeline query, not per-pipeline get; raise its default --limit 15 when needed, state the covered span, and hydrate ranked sets through one batched get.

Recipe Cards

Leadership Reliability Review / On-Call Handoff

Load references/reliability-review.md. Resolve leader, platform, and window; cluster themes, separate cause from mitigation, connect prevention work, and rank actions. For handoffs, add a first-hour checklist and escalation map.

Incident Investigation / Mitigation

Use for active, newly mitigated, or pre-PIR incidents. Anchor on the incident, then gather responders, symptoms, impact, recent deploys/flags/config, topology, alert/log/metric pointers, ownership, runbooks, and similar cases. If fields are sparse, probe four golden-signal families with bounded follow-ups. See references/incident-investigation.md. Output a four-signal matrix, hypotheses, confidence, next checks, and mitigation options; never call mitigation the root cause without its mechanism.

Post-Incident Root Cause / Learning

Use after mitigation/recovery when drafting or evaluating a PIR. Pair the incident with the PIR, linked docs, final comms, remediation PRs, and actions; cover mitigation, causal mechanism, 5-whys, and detection/response gaps. See references/pir-root-cause.md. Output root cause, contributing factors, mitigation-versus-cause, and prioritized actions.

Assets / Asset Refresh

Build contributors from project/goal/Jira/PR/doc/activity evidence. Inspect Assets schema/type metadata before AQL; join people through discovered user-like attributes such as Calculated user. Rank contribution centrality plus asset risk and report confidence/gaps. See references/assets.md.

Capacity / Staffing / Meetings

For staffing, resolve project/topic/org and identify people by related work, ownership, review influence, docs, and project/goal involvement; check load before recommending. For meetings, query scoped recordings, preview transcripts, fetch full transcripts only for central ones, then summarize decisions, actions, and gaps.

Output Shape

  • Lead with severity, urgency, or recommendation, then owner, status, recency, impact, confidence, and evidence.
  • For active investigations, add a four-signal matrix and an incident-to- learning timeline covering problem, mitigation, root-cause status, and prevention.
  • Group artifact patterns, rank next actions with a suggested owner, and call out gaps such as missing transcripts, ACL/auth failures, stale updates, or weak ownership.

Anti-Patterns

  • Do not fetch every transcript/page body or treat every incident mention as a live risk.
  • Do not wait for chat/comments to surface directional pre-PIR hypotheses.
  • Do not call mitigation root cause without a mechanism or treat workflow panels, bot comments, or opaque fields as RCA narrative.
  • Do not use keyword matches alone as ownership; cross-check assignee, service owner, PIR participants, or org-tree membership.
  • Do not join Assets by display-name guesses or recommend staffing from activity counts alone.

References

  • references/assets.md - schema-first Assets queries and person/device joins
  • references/reliability-review.md - bounded incident/PIR leadership review
  • references/incident-investigation.md - active investigation and mitigation
  • references/pir-root-cause.md - post-mitigation root-cause and PIR workflow

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.