aiops: claude - Implement repository-specific Claude skills #262

AI Ops Summary

Implement reusable Claude skills for lightspeedwp/ThisOneTimeOnAcidCom.

Build Summary

Delivery and remaining validation — 17 October 2026

Skill definitions and supporting material were delivered in aiops: claude - Configure agents, hooks, skills and cloud settings, merged into develop. Discovery, invocation, arguments, metadata, referenced paths, tool restrictions, approval boundaries and overlap with agents remain to be validated. Acceptance checkboxes remain unchanged; author-reported tests are not independently verified.

Identify reusable workflows and implement suitable .claude/skills/<skill>/SKILL.md definitions with supporting references. Keep executable helpers under scripts/claude/.

Linked Stories/Tasks/PRs/Epic

Parent: #218. Dependencies: #259, #260; hook integration where applicable: #261.

Milestones & Timeline

No milestone or deadline specified. Estimates remain unset.

Problem / Opportunity

Reusable workflows need explicit invocation, outputs and approval boundaries without overlapping agent responsibilities.

Approach / Solution

Follow repository instructions, agree agent responsibilities and verify supported skill metadata before implementation.

Acceptance Criteria

• AI workflow/agent described
• Problem/opportunity scoped
• Success metric defined
• Documentation updated
• PR uses correct branch prefix (aiops/)
• PR description updated with relevant details
• Changelog entry prepared for PR (if applicable)
• Labels/types match org standards
• Each skill has a clear purpose, invocation behaviour, inputs and expected outputs.
• Supported metadata, referenced paths and tool restrictions are validated.
• Responsibilities do not conflict with agent definitions.
• Manual invocation, intended automatic invocation and argument handling are tested.
• Side effects, approval boundaries and maintenance responsibility are documented.
• Unsupported workflow features have documented fallbacks rather than invented configuration.

Additional Context

Recheck branch-repair and rollback instructions for destructive operations. Preserve unmerged commits and unrelated changes; reconcile skill inventories, references and claimed checks with implementation.

Dependencies: settings foundations and agreed agent responsibilities. Hook integration depends on the hooks sub-issue where applicable. Branch-prefix policy requires repository verification. Preserve protected files; no CHANGELOG.md edits without separate authorisation.

Definition of Ready (DoR)

• AI ops goal described
• Area/action mapped
• Acceptance criteria listed
• Estimate added

Definition of Done (DoD)

• All acceptance criteria met
• Solution/automation verified
• Documentation updated
• PR uses correct branch prefix (aiops/)
• PR uses the correct PR template based on type:aiops PR label or linked issue type or linked issue type:aiops label
• Branch deleted after merge
• Linked issue(s) updated with latest status and closed after merge,
• The related epic should not be closed, instead updated with a comment to reflect the closed issue

Branch instructions

• Repository: lightspeedwp/ThisOneTimeOnAcidCom.
• Implementing target: this repository-skill issue, independently verified counterpart GitHub issue 262. Five delivered skills still require metadata, references, invocation, arguments, side-effect and agent-overlap validation.
• Parent context: coordination only; retain hierarchy. Omit parent identifiers/URLs from branch names, PR metadata and coding prompts while association effects remain unconfirmed. Never use the parent as a native PR development or closing target.
• Allowed closing targets: none currently; whole-outcome acceptance remains unverified.
• Branch state: proposed new branch; not created. No suitable existing implementation branch verified for reuse.
• Branch: aiops/claude-skill-boundaries.
• Base: develop, checked at 4ecdb26d9655a48bf000a7c89da2f70383edce87.
• Classification: reusable AI workflow behaviour and approval boundaries; classify independently from the Linear label.
• Naming policy: local guide, validator and organisation strategy. aiops is authorised; proposed name checked statically only.
• PR template: verified shared AI-ops template; native inheritance/named-template loading unresolved. Do not add local templates, copy sample closing/parent references or apply template-seeded labels automatically.
• Existing work: historical configuration PR is merged; open governance inventories do not establish a matching implementation branch. Neither is automatically reused.
• Missing evidence: installed-client support and operational acceptance. Agent responsibility decisions precede overlap decisions. Coordinate shared configuration documentation serially with agents and settings/hook work.
• Constraints: preserve unrelated work. No branch creation, reuse, rename, destructive repair, force-push, merge, deployment, installation or validation execution authorised.

Expected file changes

Provisional plan. All existing paths below were read at 4ecdb26d9655a48bf000a7c89da2f70383edce87. Intended edits require separate exact-content approval; path existence is not completion evidence.

No helper/test paths invented: executable remediation requires fresh inspection and a revised bounded plan. No moves or deletions proposed. Settings, hooks, cloud, application files, governance inventories and protected/locked configuration remain outside scope. Centrally managed workflows and shared locked templates/label/type definitions are not agent-edit targets.