Workflow Research Process

Research Process

Act as research partner with broad expertise spanning technical, product, business, and market domains. Your role is learning, exploration, and discovery.

Purpose in the Workflow

The exploration phase, entered from discovery — explore feasibility (technical, business, market), validate assumptions, and document findings before discussion begins.

What This Skill Needs

• Topic (required) - What to research/explore
• Output path (required) - Research file path from the handoff
• Work type (required) - epic, feature, or cross-cutting. Determines session behaviour — epic sessions carry topic awareness and reroute a grown thread to its own topic; feature and cross-cutting use the single-topic session
• Context (optional) - Prior research, constraints, starting direction

Instructions

Load framework.md and follow its instructions as written.

Resuming After Context Refresh

Context refresh (compaction) summarizes the conversation, losing procedural detail. When you detect a context refresh has occurred — the conversation feels abruptly shorter, you lack memory of recent steps, or a summary precedes this message — follow this recovery protocol:

1. Re-read this skill file completely, then re-load framework.md. Do not rely on your summary of either, and re-read both even if you believe they are already loaded — that belief is what a summary feels like from the inside. The full process, steps, and rules must be reloaded.
2. Read all research files in .workflows/{work_unit}/research/. These are the working documents this skill creates. Their content is your source of truth for progress.
3. Check agent state. Run node .claude/skills/workflow-engine/scripts/engine.cjs agent scan {work_unit} research {topic} — in_flight agents still running, pending results unread, acknowledged results partially surfaced.
4. Check git state. Run git status and git log --oneline -10 to see recent commits. Commit messages follow a conventional pattern that reveals what was completed.
5. Announce your position to the user before continuing: what step you believe you're at, what's been completed, and what comes next. Wait for confirmation.

Do not guess at progress or continue from memory. The files on disk and git history are authoritative — your recollection is not.

Step 0: Resume Detection

Refresh the tmux session label — a no-op unless the user opted in and this session runs inside tmux:

node .claude/skills/workflow-engine/scripts/engine.cjs session label {work_unit} research {topic}

Read the phase status:

node .claude/skills/workflow-engine/scripts/engine.cjs manifest get {work_unit}.research.{topic} status

Then check if the research file exists at .workflows/{work_unit}/research/{topic}.md.

If status is triaged

A first start, not a resume — no session has ever run. Parked concerns wait in the topic's triage queue, untouched by initialization — the session loop's triage check surfaces them.

→ Proceed to Step 1.

If no file exists

→ Proceed to Step 1.

Otherwise

Output the next fenced block as markdown (not a code block):

**`□ Resume Detection`**

Output the next fenced block as markdown (not a code block):

> An in-progress research file exists for this topic — choose whether to pick it up or start fresh.

Load resume-detection.md with artifact = research, file = .workflows/{work_unit}/research/{topic}.md, continue_step = Step 2, restart_targets = the research file and the phase cache directory (rm -rf .workflows/.cache/{work_unit}/research/{topic}/ — content and agent state together) — stale agent results would poison the restarted session's review gates, commit = research({work_unit}): restart research.

Step 1: Initialize Research

Load initialize-research.md and follow its instructions as written.

→ On return, proceed to Step 2.

Step 2: File Strategy

Load file-strategy.md and follow its instructions as written.

→ On return, proceed to Step 3.

Step 3: Research Guidelines

Load research-guidelines.md and follow its instructions as written.

→ On return, proceed to Step 4.

Step 4: Knowledge Usage

Load knowledge-usage.md and follow its instructions as written.

→ On return, proceed to Step 5.

Step 5: Contextual Query

Load contextual-query.md and follow its instructions as written.

→ On return, proceed to Step 6.

Step 6: Research Session

Output the next fenced block as markdown (not a code block):

**`□ Research Session`**

Output the next fenced block as markdown (not a code block):

> Starting the research session. This is open-ended exploration — follow threads, surface options, and document findings. No decisions needed at this stage.

Load route-session.md and follow its instructions as written.

*Knowledge-base nudge — if a thread feels familiar, or you're about to re-tread ground that might have been covered in another work unit, run a quick query before proceeding. See knowledge-usage.md.*