trace-first-starter

RF — TFW-24 / Phase B: Research Stage Templates

Date: 2026-04-04 Author: Executor Status: 🟢 RF — Complete Parent HL: HL-TFW-24 TS: TS Phase B

---

1. What Was Done

New Files

| File | Description | |——|————| | .tfw/templates/research/briefing.md | Briefing template: Research Plan, Hypotheses, Scope Intent, Guiding Questions, User Direction | | .tfw/templates/research/gather.md | Gather stage: “What do we NOT know?” — Findings, Checkpoint, Sufficiency | | .tfw/templates/research/extract.md | Extract stage: “What do we NOT see?” — Findings, Checkpoint, Sufficiency | | .tfw/templates/research/challenge.md | Challenge stage: “What do we NOT expect?” — Findings, Checkpoint, Sufficiency |

Modified Files

| File | Changes | |——|———| | .tfw/workflows/research/base.md | Step 3: templates/research/ ref. Step 4: templates/research/briefing.md ref. Step 5: template ref | | .tfw/conventions.md | §4: replaced 18-line inline format block with 1-line template reference | | .agent/workflows/tfw-research.md | Synced from base.md |

2. Key Decisions

1. Step 5 wording compressed. TS specified “Each stage uses its template from templates/research/ and writes to research/” — this pushed base.md to 604 words. Trimmed to “Each stage uses its template (templates/research/)” — 599 words. “writes to research/” is implied by the template ref and Step 3’s subfolder creation.

3. Acceptance Criteria

• 4 template files exist in .tfw/templates/research/
• Each stage template has: Parent HL link, Goal from §1 Vision, Findings section, Checkpoint with Stage complete: YES / NO
• Each stage template has its guiding question as subtitle (D28): Gather = “What do we NOT know?”, Extract = “What do we NOT see?”, Challenge = “What do we NOT expect?”
• Briefing template includes: Research Plan, Hypotheses table, Scope Intent, Guiding Questions, User Direction
• Gather/Extract/Challenge templates include Sufficiency checklist (external source + briefing gap)
• base.md Step 3 references templates/research/
• base.md Step 4 references templates/research/briefing.md
• base.md Step 5 references templates/research/ for stage files
• conventions.md §4 references templates instead of inline format
• Adapter synced (diff = 0)

4. Verification

• Lint: N/A (markdown-only project)
• Tests: N/A
• Verify: Passed
• wc -w base.md: 599 words ✅ (under 600 budget)
• diff base.md adapter: 0 differences ✅
• Template count: 4 files in .tfw/templates/research/ ✅

Deviation from TS

#	TS spec	Actual	Reason
1	Step 5: “…and writes to research/”	“Each stage uses its template (templates/research/)”	Word count: TS wording = 604 words (over budget). Trimmed redundant clause — writing target implied by Step 3 subfolder creation

5. Observations (out-of-scope, not modified)

No observations.

6. Fact Candidates

#	Category	Candidate	Source	Confidence
1	convention	Research stage templates live in .tfw/templates/research/ (briefing, gather, extract, challenge). These are starting points, not contracts — Findings section is free-form. Canonical format for Goal anchoring and Checkpoint consistency	TS Phase B, this RF	High

---

*RF — TFW-24 / Phase B: Research Stage Templates

2026-04-04*

This site is open source. Improve this page.