Drop SR-N identifiers from closeout artifacts and SREQ template #4

New issue

Open

opened 2026-04-20 19:28:27 +00:00 by jbr870 · 0 comments

jbr870 commented

2026-04-20 19:28:27 +00:00

Owner

Problem

After #3 fixed the sreqs/SR-*/ directory path drift, residual SR-N naming still appears in consumer artifact templates and in the SREQ template itself, even though the current model has:

One SREQ.md per feature (no SR-N decomposition at the SREQ layer)
Work-unit (WU-N) decomposition happens inside /develop at runtime
Commits are prefixed with feat($FEATURE_SLUG):, not feat(SR-001): (see .claude/skills/develop/SKILL.md:151)

These remaining sites don't break path resolution the way #3 did — they shape the downstream generated documents (traceability matrix, new SREQ scaffolds) around an entity (SR-N) that no longer exists.

Drift sites

`.claude/skills/closeout/procedures/generate-docs.md`

Line	Current	Problem
30	`api: abc1234 feat(SR-001): ...`	Commits never use `SR-N` prefix — `/develop` emits `feat($FEATURE_SLUG):`
51–54	Traceability matrix has `SREQ(s)` column with `SR-001`, `SR-002, SR-003` entries	Should map ACs to `WU-N` work units
56–60	Separate `SREQ Summary` table (`SR-001 \| title \| tasks \| commits \| status`)	No `SR-N` entities exist — would be a `Work Units Summary`
80	`[List all files created/modified across all SREQs]`	Should be "across all work units"
91	`SREQs table — populate with SR-XXX IDs, titles, statuses`	Refers to a PREQ.md section that doesn't exist in `default-preq-template.md`

`.claude/skills/_shared/templates/default-sreq-template.md`

L1 — # SR-XXX: <Title> — vestigial; when instantiated as $FEATURE_FOLDER/SREQ.md there's no SR-N to disambiguate
L5 — **Dependencies:** [SR-YYY or none] — no sibling SREQs to depend on in the single-file model

Minor stragglers

closeout/procedures/learning-analysis.md:50 — example "SR-002 was estimated at 2 hours but took 5..." — should use WU-N
integrate/SKILL.md:98 — "no SREQs to trace against" — minor wording, reads as plural/multi

Proposed fix

Redesign the traceability matrix template in generate-docs.md: replace the SREQ(s) column with Work Units (WU-N values come from /develop's decomposition), drop the SREQ Summary table in favor of a Work Units Summary populated from /develop's final report.
Fix the commit-example prefix to feat($FEATURE_SLUG):.
Drop the "SREQs table" reference in generate-docs.md step 3 (no such table in PREQ).
Reshape default-sreq-template.md as a single-SREQ document — header like # SREQ: <Feature Name>, drop the SR-YYY Dependencies field.
Update example wording in learning-analysis.md:50 and integrate/SKILL.md:98.

Why this wasn't bundled with #3

#3's scope was path mismatches — consumers globbing a directory that no longer exists. The residual drift here modifies the shape of downstream generated artifacts (what a traceability matrix looks like, what a new SREQ scaffold looks like). The matrix redesign in particular is worth discussing before landing — deciding whether ACs map 1:1 to WU-N or whether WU-N should even appear in the external MR/PR deliverable.

#3 — sreqs/SR-*/ path drift (fixed in commits 4ff8a84, d34aeea)
#1 — YAML frontmatter for journals
#2 — consolidate drifted journal templates

## Problem After #3 fixed the `sreqs/SR-*/` directory path drift, residual SR-N naming still appears in consumer **artifact templates** and in the SREQ template itself, even though the current model has: - One `SREQ.md` per feature (no SR-N decomposition at the SREQ layer) - Work-unit (WU-N) decomposition happens inside `/develop` at runtime - Commits are prefixed with `feat($FEATURE_SLUG):`, not `feat(SR-001):` (see `.claude/skills/develop/SKILL.md:151`) These remaining sites don't break path resolution the way #3 did — they shape the **downstream generated documents** (traceability matrix, new SREQ scaffolds) around an entity (`SR-N`) that no longer exists. ## Drift sites ### `.claude/skills/closeout/procedures/generate-docs.md` | Line | Current | Problem | |---|---|---| | 30 | `` `api: abc1234 feat(SR-001): ...` `` | Commits never use `SR-N` prefix — `/develop` emits `feat($FEATURE_SLUG):` | | 51–54 | Traceability matrix has `SREQ(s)` column with `SR-001`, `SR-002, SR-003` entries | Should map ACs to `WU-N` work units | | 56–60 | Separate `SREQ Summary` table (`SR-001 \| title \| tasks \| commits \| status`) | No `SR-N` entities exist — would be a `Work Units Summary` | | 80 | `[List all files created/modified across all SREQs]` | Should be "across all work units" | | 91 | `**SREQs table** — populate with SR-XXX IDs, titles, statuses` | Refers to a PREQ.md section that doesn't exist in `default-preq-template.md` | ### `.claude/skills/_shared/templates/default-sreq-template.md` - **L1** — `# SR-XXX: <Title>` — vestigial; when instantiated as `$FEATURE_FOLDER/SREQ.md` there's no `SR-N` to disambiguate - **L5** — `**Dependencies:** [SR-YYY or none]` — no sibling SREQs to depend on in the single-file model ### Minor stragglers - `closeout/procedures/learning-analysis.md:50` — example ``"SR-002 was estimated at 2 hours but took 5..."`` — should use `WU-N` - `integrate/SKILL.md:98` — "no SREQs to trace against" — minor wording, reads as plural/multi ## Proposed fix 1. Redesign the traceability matrix template in `generate-docs.md`: replace the `SREQ(s)` column with `Work Units` (WU-N values come from /develop's decomposition), drop the `SREQ Summary` table in favor of a `Work Units Summary` populated from /develop's final report. 2. Fix the commit-example prefix to `feat($FEATURE_SLUG):`. 3. Drop the "SREQs table" reference in generate-docs.md step 3 (no such table in PREQ). 4. Reshape `default-sreq-template.md` as a single-SREQ document — header like `# SREQ: <Feature Name>`, drop the SR-YYY Dependencies field. 5. Update example wording in `learning-analysis.md:50` and `integrate/SKILL.md:98`. ## Why this wasn't bundled with #3 #3's scope was **path mismatches** — consumers globbing a directory that no longer exists. The residual drift here modifies the **shape of downstream generated artifacts** (what a traceability matrix looks like, what a new SREQ scaffold looks like). The matrix redesign in particular is worth discussing before landing — deciding whether ACs map 1:1 to WU-N or whether WU-N should even appear in the external MR/PR deliverable. ## Related - #3 — `sreqs/SR-*/` path drift (fixed in commits `4ff8a84`, `d34aeea`) - #1 — YAML frontmatter for journals - #2 — consolidate drifted journal templates

No labels

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference

jbr870/devwork-skills#4

No description provided.

Rows
Columns