Bryan Mendoza

Run the conversational clarification layer for a workflow orchestrator. `MODE=upfront` is the plan-wide clarification pass and `MODE=critique` is the task-level pre-execution clarification pass. Keeps the mentoring dialogue inline while delegating artifact reading, critique generation, manifest assembly, and file updates to subagents.

Phase 4 skill for clarified GitHub task plans. Use after the task plan has been clarified and the user has explicitly approved GitHub writes. This skill reads only its bundled files, dispatches the `task-issue-creator` subagent with `ISSUE_URL`, and returns a concise phase summary after GitHub task issues are created or reconciled and `docs/<ISSUE_SLUG>-tasks.md` is updated with the Phase 4 output contract.

Phase 4 skill for clarified Jira task plans. Use after the task plan has been clarified and the user has explicitly approved Jira writes. This skill reads only its bundled files, dispatches the `subtask-creator` subagent with `JIRA_URL`, and returns a concise phase summary after Jira subtasks are created or reconciled and `docs/<TICKET_KEY>-tasks.md` is updated with the Phase 4 output contract.

Execute exactly one planned GitHub workflow task using pre-produced task artifacts and a specialist pipeline. Use when the user says "execute task 2", "implement task 4", or "work on task 1 for acme-app-42". Requires the issue snapshot, task plan, per-task planning artifacts, critique record, and decisions record for the selected task. Execution begins with an explicit kickoff, the first mutation boundary after critique approval, then continues through implementation, documentation, requirements verification, review gates, targeted fix cycles, and final reporting for one task only.

Execute exactly one planned Jira workflow task using pre-produced task artifacts and a specialist pipeline. Use when the user says "execute task 2", "implement task 4", or "work on task 1 for PROJ-123". Requires the task snapshot, task plan, per-task planning artifacts, critique record, and decisions record for the selected task. Execution begins with an explicit kickoff, the first mutation boundary after critique approval, then continues through implementation, documentation, requirements verification, review gates, targeted fix cycles, and final reporting for one task only.

Phase 1 of `orchestrating-github-workflow`: retrieve a GitHub issue into a stable Markdown snapshot for downstream workflow phases. Use this as a workflow phase, not as a standalone implementation skill, when an issue URL or owner/repo/number needs to become `docs/<ISSUE_SLUG>.md` with predictable headings for metadata, description, acceptance criteria, comments, child issues, linked issues, labels, assignees, and optional milestone/projects/attachments. The bundled retriever handles GitHub reads, validation, and snapshot assembly. This skill coordinates retrieval only: it does not modify GitHub state beyond read-only queries, create branches, or start implementation.

Phase 1 of `orchestrating-jira-workflow`: retrieve a Jira ticket into a stable Markdown snapshot for downstream workflow phases. Use this as a workflow phase, not as a standalone implementation skill, whenever a Jira URL needs to become `docs/<TICKET_KEY>.md` with predictable headings for metadata, description, acceptance criteria, comments, subtasks, linked issues, attachments, and custom fields. The bundled retriever handles Jira reads, validation, and snapshot assembly. This skill coordinates retrieval only: it does not modify Jira, create branches, or start implementation.

Generate a resumable handoff document from an in-progress conversation, review, debugging session, or investigation. Dispatches co-located subagents to extract original instructions and Q&A context, capture evidence-backed insights, optionally validate claims from tracking files, and assemble a cold-start-ready handoff file plus structured working artifacts. Use when the user says "create a handoff doc", "save this for later", "document what we found", "update the resumption file", or wants a fresh agent to resume later without relying on chat history.

Coordinate an end-to-end GitHub issue workflow from issue fetch through per-task implementation. Use this skill when the user provides a GitHub issue URL, says "work on issue owner/repo#123", "resume <issue-slug>", "continue this GitHub issue", "start the GitHub workflow", or asks for status on an issue without naming a specific phase. This skill is the top-level coordinator: it reads only the skill/reference files it needs, talks to the user, and dispatches all execution-heavy work to downstream skills and co-located utility subagents. Primary GitHub transport for delegated work is gh (GitHub CLI).

Coordinate an end-to-end Jira ticket workflow from ticket fetch through per-task implementation. Use this skill when the user provides a Jira URL, says "work on ticket PROJECT-123", "resume PROJECT-123", "continue this Jira ticket", "start the Jira workflow", or asks for status on a ticket without naming a specific phase. This skill is the top-level coordinator: it reads only the skill/reference files it needs, talks to the user, and dispatches all execution-heavy work to downstream skills and co-located utility subagents.

Phase 2 of the GitHub planning workflow. Reads a GitHub issue snapshot at docs/<ISSUE_SLUG>.md and produces a detailed, self-contained task plan at docs/<ISSUE_SLUG>-tasks.md through a three-stage subagent pipeline (plan → prioritize → validate). Preserves planning artifacts for resume and critique and returns only concise handoff summaries.

Plan how to execute one task from `docs/<ISSUE_SLUG>-tasks.md`. Runs a four-subagent pipeline that validates the task, writes an execution brief, inspects the codebase, defines behavior-driven tests, and evaluates refactoring needs. Produces `docs/<ISSUE_SLUG>-task-<TASK_NUMBER>-{brief,execution-plan,test-spec,refactoring-plan}.md` for one task only. Use when the user says "plan task 3", "prepare task 2 for execution", "how should we implement task 1", or when invoked as the planning phase of a multi-phase workflow. This skill creates planning artifacts only; it does not change git state, mutate GitHub issues, or modify product code.

Plan how to execute one task from `docs/<TICKET_KEY>-tasks.md`. Runs a four-subagent pipeline that validates the task, writes an execution brief, inspects the codebase, defines behavior-driven tests, and evaluates refactoring needs. Produces `docs/<TICKET_KEY>-task-<TASK_NUMBER>-{brief,execution-plan,test-spec,refactoring-plan}.md` for one task only. Use when the user says "plan task 3", "prepare task 2 for execution", "how should we implement task 1", or when invoked as the planning phase of a multi-phase workflow. This skill creates planning artifacts only; it does not change git state, transition Jira issues, or modify product code.

Phase 2 of the Jira planning workflow. Reads a Jira ticket snapshot at docs/<TICKET_KEY>.md and produces a detailed, self-contained task plan at docs/<TICKET_KEY>-tasks.md through a three-stage subagent pipeline (plan → prioritize → validate). Preserves planning artifacts for resume and critique and returns only concise handoff summaries.

Create review-ready pull requests from the current branch by validating repo state, comparing against a user-specified base branch, drafting a title and body from the real branch diff, suggesting reviewers and labels, previewing the result, and creating the PR only after explicit confirmation. Use when the user asks to create or open a PR, draft pull request, merge request, code review request, or says their work is ready for review.

Convert prose prompts into structured XML prompts using a five-pass methodology. Use this skill whenever a user wants to turn a text-based prompt, instruction block, or natural-language request into a more structured, tagged, agent-friendly format — including phrases like 'structure this prompt', 'make this prompt more reliable', 'convert this to XML', 'formalize this prompt', 'this prompt keeps failing, can you tighten it up', or 'turn this into a proper prompt template'. Also use when a user provides a prompt that has ambiguity, implicit assumptions, or agent-drift problems and wants it reshaped. Works for single prompts and prompt suites (multiple related prompts that should stay internally consistent).

Validate answers whose value depends on current external facts. Use this skill whenever a response includes time-sensitive claims, rankings, recommendations, pricing, version status, policy changes, availability, or other facts that could have changed recently. Also use it when the user asks for a verified, fact-checked, or up-to-date answer. This skill runs a sequential validation pipeline: recency audit, claim stress-test, completeness check, and clarity pass. Skip it for purely creative writing, casual chat, or requests with no factual claims to verify.

Audit an AI-generated implementation plan safely. Use when reviewing or validating a plan or design proposal and you want a standalone audit report for requirements traceability, YAGNI compliance, and risky assumptions without loading the raw plan into the orchestrator or overwriting the source file.

Convert multi-step workflows into production-ready Claude Code skills and subagents. Use this skill whenever a user describes a workflow, process, or multi-step procedure and wants to turn it (or any part of it) into reusable Claude Code skills, subagents, or slash commands. Also trigger when the user says things like "make this a skill", "turn this process into an agent", "I want to automate this workflow", "break this into skills", "create skills from these steps", or asks how to structure a set of related tasks as Claude Code artifacts. Works for any domain — DevOps pipelines, content creation, data processing, customer support, research, onboarding, code review, or any repeatable process.