dotfiles

Read, create, update, inspect, and reply to Bitbucket pull requests or commit diffs using Bitbucket URLs plus environment-backed auth. Use when Codex needs Bitbucket API operations for PR metadata, compare or commit evidence, PR discussion threads, PR creation/update, reviewer preservation, or prefixed replies. This skill does not perform full code-change review or own code remediation.

Review change-management tickets such as Jira CHANGE or CM requests for deployment readiness, scope drift, documentation quality, and runbook adherence. Use when Codex needs to assess implementation steps, validation evidence, rollback clarity, Shepherd release scope and plan diffs, regional outliers, manual data-fix or host-maintenance procedures, team-specific runbook alignment, and commit or artifact-version alignment, and optionally write findings back to the ticket.

Bootstrap a local Codex setup from this shared starter repo. Use when an engineer needs an interactive onboarding flow that asks for local paths, required MCP settings, and only the overrides that differ from the shared Enterprise ChatGPT defaults, then generates or repairs a real local Codex config and verifies that shared path placeholders were resolved.

Use when onboarding OCI users to available packs, registry entries, PromptLib spaces, ORA skills, Skills Hub / AI Skills Registry, Codex Enterprise apps/connectors, or non-pack content sources after installing OCI starter packs.

Inspect DevOps MFO or region-build flock dependency status for a specific region, project, flock, and phonebookId. Use when Codex needs to determine which region-build phases are satisfied, unsatisfied, or optional; identify whether a flock has published the required capabilities; recursively trace missing capability producers; discover related infrastructure or application flocks from the dependency graph; or explain which upstream project or flock is blocking publication.

Preflights local ossh and oci-ops access, reads a Jira ticket to identify a target hostname and region, validates OCI authentication, resolves the tenancy OCID, and returns the appropriate oci-ops SSH connection command. Use when a user asks to SSH to an OCI host from a Jira ticket, set up ossh or oci-ops access, resolve a host region or tenancy, or needs an oci-ops SSH command.

Use when running production incident triage for an on-call service and you need a repeatable, config-driven workflow that correlates ticket evidence, metrics, logs, releases, and code before writing back findings.

Inspect Oracle Ticketing Platform tickets and related resources from an OCI CLI session-backed profile. Use when Codex needs to fetch ticket details, comments, activity history, project metadata, TQL search results, dashboards, subQueries, linked tickets, attachment resources, post ticket comments to ticketing-platform.us-chicago-1.oci.oraclecloud.com, or mutate ticket labels.

Draft pull request descriptions from the current branch using the shared PR template and branch evidence. Use when Codex needs to prepare a PR body, summarize branch changes for reviewers, fill the checklist from actual evidence, or recommend the final commit message text.

Use when a Codex skill declares a remote SKILL.md source and needs to warn whether the active local skill version is stale or unverifiable before continuing.

SCM PR Operations for OCI DevOps pull requests from a local checkout or known repository OCID. This is the canonical shared skill for PR discovery, PR creation, comment inspection, threaded replies, Codex-origin freeform tagging, and review-complete freeform tagging; repo-local wrappers can layer repository-specific defaults on top.

Set up or reactivate service OKE access for a given OCI realm. Use when the user wants service JIT activation and propagation preflight, session-auth commands, SSH tunnel commands, OKE kubectl verification, service SSH config entries, OKE/database tunnels, or per-realm kubeconfig backup files after the user completes external-browser OCI tenancy authentication. Derive region from local realm state and use a realm-derived profile name.

troubleshoot redwood migration engineering questions using oracle central confluence as the source of truth. use when a user asks about maui template migration, config authoring, v2 vs v3 templates, list/details/forms/settings templates, selection/row id issues, event payloads, or any redwood migration bug and they provide a problem description, config snippet, logs, or screenshots.

Create or update implementation plans for the console-ui-testing-agent repository. Use when Codex is asked to write an implementation plan, planning doc, ticket plan, or scoped work plan for this repo; plans must live under docs/implementation-plans and the filename must include the ticket number, asking the user for the ticket number when it is not known.

Generate authz resourceOperations YAML for OCI services from SPLAT spec permissions and permissions documentation. Use when creating or updating ID-*.yaml permission files and resource families.

Use when a repository needs module-specific Codex skills, repo AGENTS.md skill routing, or drift-maintained knowledge packs for codebase modules, architecture areas, services, packages, or Maven/Gradle/workspace submodules.

Use when setting up MCP servers from a cold start, onboarding a new environment, or troubleshooting a server that won't start — covers prereqs, install patterns, auth, and verification

Extract, authenticate, and summarize internal Confluence pages by rendering the visible page content to Markdown first and then working from the saved file. Use when a user shares an internal Confluence URL and asks for a summary, technical review, or other text analysis that requires SSO, cookie refresh, or reproducible output files.

Read and write Jira-SD or Jira OCI tickets using the shared Atlassian MCP env files or explicit CLI overrides. Use when Codex needs to build JQL from filters such as project, severity, ticket key, time window, or sorting; search Jira; read issue details with custom fields; inspect comments and activity history; add and reply to issue comments; or mutate Jira labels.

Use this skill when you need to access Oracle Object Storage from this repo, especially to resolve a tenancy or account, discover the namespace, list buckets or objects, download logs, or search downloaded files. Prefer direct `oci` CLI commands on the local machine, including `oci raw-request` for SPLAT account lookup, and reuse the shared on-call team config when a team has stable tenancy, namespace, bucket, or prefix conventions.

Use when setting up a new OCA harness from scratch, resolving OCA-specific issues (model prefix, variant config paths), or comparing harness capabilities — not for adding rules/skills/MCP to an existing harness (use agent-configuration for that)

Use when orienting OCI users to AI work at Oracle/OCI/Platform, troubleshooting first-run setup, or deciding where to search next across AIPack, Codex/OCA, MCP Gateway, PromptLib, Confluence, Slack, apps/connectors, and team packs.

Review or investigate OCI Shepherd release links for rollout scope, phase and target status, expected and observed diffs, per-region changes, execution errors, target logs, validations, timelines, and follow-up recommendations. Use when Codex is given a Shepherd release URL or release identifiers and needs to confirm whether a deployment looks correct, analyze what a release will change, compare target-region diffs, inspect release or target logs, identify failing execution targets or release targets, or summarize risks and next steps.

Use when understanding, adding, changing, or troubleshooting agent harness configuration (rules/skills/commands/tools/permissions) across Claude Code, OpenCode, Codex, or Cline

Use when syncing, configuring, troubleshooting, or managing aipack packs — including sync-config, profiles, harness behaviors, and the delivery pipeline

Use when you need to write a Grafana annotation for a confirmed Autoschrute incident window after the mutation gate has been explicitly approved.

[Workflow] Investigate elevated 5xx windows for an OCM service and optionally write a Grafana annotation after explicit approval.

[Agent] Confluence domain specialist — navigates OCM/OCB spaces using CQL, assesses page freshness and ownership, maps page relationships, returns ranked results with confidence signals

Use when reviewing a design document for ECAR, PreCAR, or equivalent OCI architecture review and you need to assess applicability, required sections, SDP coverage, action items, and readiness before producing feedback.

Use when you need to scan a date range for elevated 5xx windows for a specific region, T2 project, and fleet using the packaged parser script.

[Workflow] Create a concise handoff note for the next oncall using jira tickets

Use when maintaining OCM engineering workflows across multiple repos or services and you need the canonical OCM runbooks, shepherd/release anchors, prerequisites context, or TC handoff guidance.

[Workflow] Execution-first docs drift check: map product/ops changes to required documentation updates with evidence

[Workflow] Execution-first postmortem workflow (timeline, impact, fixes, action items) with evidence requirements and mutation gating

[Workflow] Execution-first workflow for triaging stuck/failed OCM migrations (read-only by default)

[Workflow] Investigate OCM service symptoms using dashboards, logs, and metrics.

Use when triaging OCM/OCB incidents such as 5xx spikes, OTS tickets, or customer impact and you need OCM-specific mappings, telemetry context, and handoff queries.

[Workflow] Check OCM prerequisites stack drift (v2.3)

[Workflow] Execution-first OCM release readiness checklist (read-only by default)

[Workflow] Execution workflow for OCM Shepherd deployments (read-only by default; explicit mutation gate)