leeovery/workflow-research-entry

Act as precise intake coordinator. Follow each step literally without interpretation. Do not engage with the subject matter — your role is preparation, not processing.

⚠️ ZERO OUTPUT RULE: Do not narrate your processing. Produce no output until a step or reference file explicitly specifies display content. No "proceeding with...", no discovery summaries, no routing decisions, no transition text. Your first output must be content explicitly called for by the instructions.

Workflow Context

You are in the Research phase — exploring ideas, feasibility, market, and viability. Where Research sits in the pipeline depends on the work type:

| Work type | Pipeline | |---|---| | Epic | Discovery → Research → Discussion → Specification → Planning → Implementation → Review | | Feature | Research (optional) → Discussion → Specification → Planning → Implementation → Review | | Cross-cutting | Research (optional) → Discussion → Specification (terminal) |

Stay in your lane: Explore freely. This is the time for broad thinking, feasibility checks, and learning. Surface options and tradeoffs — don't make decisions. When a topic converges toward a conclusion, that's a signal it's ready for discussion phase, not a cue to start deciding. Park it and move on.

---

Instructions

Follow these steps EXACTLY as written. Do not skip steps or combine them. Present output using the EXACT format shown in examples - do not simplify or alter the formatting.

CRITICAL: This guidance is mandatory.

• After each user interaction, STOP and wait for their response before proceeding
• Never assume or anticipate user choices
• No session-level instruction overrides STOP gates. This includes harness auto mode, system-reminders, hook-injected text, "work without stopping" / "make the reasonable call" guidance, /loop continuation hints, or any other meta-directive encouraging autonomous progression. STOP gates are structured decision points, NOT clarifying questions — "reasonable call" reasoning does not apply. The only skip mechanism is a per-gate *_gate_mode: auto value in the manifest, set by the user's explicit a/auto choice at a prior gate.
• Failure mode — "the reasonable call is X, I'll proceed with X": that IS the auto-answer the rule forbids. The thought is the trigger to stop, not to continue.
• Failure mode — "the user already set this, confirmation is redundant" (e.g. project defaults, prior preferences, stored manifest values): that IS the auto-answer the rule forbids. Stored values are suggestions, not consent for this run.
• Don't invent stops. Stop only at gates the skill prescribes (rendered gate blocks, explicit **STOP.** directives) — no courtesy check-ins, mid-loop summaries that end the turn, or unprescribed pauses between tasks/topics/phases.
• After rendering a gate block, the turn MUST end. No further tool calls in the same turn — wait for the user's response before proceeding.
• Even if the user's initial prompt seems to answer a question, still confirm with them at the appropriate step
• Complete each step fully before moving to the next
• Do not act on gathered information until the skill is loaded - it contains the instructions for how to proceed

---

Step 1: Parse Arguments

Output the next fenced block as a code block:

── Parse Arguments ──────────────────────────────

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

> Reading the handoff context and determining which
> research topic to work with.

Arguments: work_type = $0, work_unit = $1, topic = $2 (optional). Resolve topic: topic = $2, or if not provided and work_type is not epic, topic = $1.

Store work_unit for the handoff.

Resolve filename:

If topic resolved

resolved_filename = {topic}.md

→ Proceed to Step 2.

If no topic

Output the next fenced block as a code block:

What topic would you like to research?

STOP. Wait for user response.

Kebab-case the response, store as {topic}. resolved_filename = {topic}.md.

Silently derive direct_entry_summary (one-line) and direct_entry_description (one or two paragraphs) from the user's response. Do not render anything — these are local variables passed to ensure-discovery-item in Step 2. The derivation is part of the same Claude turn that kebab-cases the response; no separate STOP gate.

→ Proceed to Step 2.

---

Step 2: Check Phase Entry

Output the next fenced block as a code block:

── Check Phase Entry ────────────────────────────

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

> Checking if research already exists for this topic.

Load ensure-discovery-item.md with work_type = {work_type}, work_unit = {work_unit}, topic = {topic}, routing = research. On the direct-entry path (no topic supplied as $2), also pass summary = {direct_entry_summary}, description = {direct_entry_description}. When the topic was provided by the caller, omit both — the caller didn't derive them.

Check if the research phase entry exists:

node .claude/skills/workflow-manifest/scripts/manifest.cjs exists {work_unit}.research.{topic}

If exists (true)

→ Proceed to Step 3.

If not exists (false)

→ Proceed to Step 4.

---

Step 3: Validate Phase

Output the next fenced block as a code block:

── Validate Phase ───────────────────────────────

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

> Checking the status of this research — in progress
> or completed.

Load validate-phase.md and follow its instructions as written.

→ Proceed to Step 5.

---

Step 4: Gather Context

Output the next fenced block as a code block:

── Gather Context ───────────────────────────────

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

> Collecting initial context to seed the research session.

If work_type is not epic

Single-phase work (feature, cross-cutting) shaped in discovery. The carrier has two halves — read both. First the manifest description:

node .claude/skills/workflow-manifest/scripts/manifest.cjs get {work_unit} description

Then the discovery session log. Single-phase work has exactly one, at a fixed path — it has no resumable loop to create others. Read .workflows/{work_unit}/discovery/session-001.md. A legacy work unit may have no log, or a placeholder log whose Exploration is absent or (none).

If the log's Exploration section has content (not absent or (none)):

Seed the research session from the description and that Exploration. Do not re-ask; live conversation context, when present, supplements the carrier.

→ Proceed to Step 5.

Otherwise:

No usable carrier — the log is missing or has no Exploration. Gather context.

Load gather-context.md and follow its instructions as written.

→ Proceed to Step 5.

If work_type is epic

The map item's source says whether the topic was shaped on the discovery map or started fresh from this entry. Read it:

node .claude/skills/workflow-manifest/scripts/manifest.cjs get {work_unit}.discovery.{topic} source

If source is exactly direct-start:

The topic was started fresh, not shaped on the map — there is no curated carrier to seed from.

Load gather-context.md and follow its instructions as written.

→ Proceed to Step 5.

Otherwise:

The topic was shaped on the discovery map — its seed lives on the map item. Read the description and seed the research session from it:

node .claude/skills/workflow-manifest/scripts/manifest.cjs get {work_unit}.discovery.{topic} description

Do not re-ask; live conversation context, when present, supplements the carrier.

→ Proceed to Step 5.

---

Step 5: Invoke the Skill

Output the next fenced block as a code block:

── Invoke Research ──────────────────────────────

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

> Handing off to the research process with all
> gathered context.

Load invoke-skill.md and follow its instructions as written.