ensemble-requirement-status

name	ensemble-requirement-status
description	On-demand requirement satisfaction report — scans bead comments for req-verified tokens (Codex skill for /ensemble:requirement-status)
user-invocable	true
argument-hint	["trd-path-or-slug"]
model	gpt-5.1-codex

Ensemble Command: /ensemble:requirement-status

This Codex skill mirrors the Ensemble slash command /ensemble:requirement-status. Follow the workflow below, adapt to the current repository, and keep outputs structured.

Scan the root epic bead and all child task beads for req-verified: tokens written by implement-trd-beads when test tasks close. Cross-reference against the PRD REQ-NNN list to produce a requirement satisfaction table showing which requirements are verified, which are in-progress, and which are not yet started.

Workflow

Phase 1: Context Resolution

1. TRD Slug Resolution Resolve the TRD slug from $ARGUMENTS. If $ARGUMENTS is a file path: derive TRD_SLUG from filename (lowercase, replace non-alphanumeric with hyphens). If $ARGUMENTS is a slug string: use directly as TRD_SLUG. If $ARGUMENTS empty: list recent TRD-prefixed epics via br list --status=open --json and prompt user. If br list returns no TRD-prefixed epics: print 'No TRD implementations found in beads storage. Run /ensemble:implement-trd-beads first.' and EXIT.

2. Root Epic Location Find the root epic bead for this TRD. Run: br list --status=open --json Verify the output is valid JSON (starts with '['). If not: print 'ERROR: br command failed or returned non-JSON output. Is br installed and configured? Output: <first 200 chars>' and EXIT. Also run: br list --status=closed --json Verify the output is valid JSON (starts with '['). If not: print 'ERROR: br command failed or returned non-JSON output. Is br installed and configured? Output: <first 200 chars>' and EXIT. Parse combined JSON output, search for entry where title matches [trd:<TRD_SLUG>] with type epic. If found: ROOT_EPIC_ID = bead .id If not found: print 'ERROR: No root epic found for TRD slug <TRD_SLUG>. Run implement-trd-beads first.' and EXIT.

3. PRD Requirement Loading Load PRD REQ-NNN list for cross-reference. Search TRD document for PRD reference link or 'Based on PRD:' annotation. If TRD has a PRD reference line AND the file exists at that path: parse PRD for REQ-NNN IDs -> build PRD_REQUIREMENTS map. If TRD has a PRD reference line BUT the file does not exist at that path: print 'WARNING: PRD file referenced in TRD not found at . Cross-reference reporting will be incomplete. Fix the PRD path in the TRD.' Set PRD_REQUIREMENTS = null. If TRD has no PRD reference line at all: print 'NOTE: No PRD reference in TRD — cross-reference not available.' Set PRD_REQUIREMENTS = null.

Phase 2: Evidence Collection

1. Epic Comment Scan Scan root epic comments for req-verified tokens. Run: br comment list <ROOT_EPIC_ID> If br comment list returns non-JSON or exits non-zero: print 'WARNING: Could not read epic comments for <ROOT_EPIC_ID> — verified requirement data may be incomplete.' Continue with VERIFIED_REQS = {}. Parse each comment line for tokens matching: req-verified:REQ-NNN by:TRD-NNN-TEST qa: ac-proven:AC-NNN-M,AC-NNN-M Build VERIFIED_REQS map: REQ-NNN -> {test_task, qa_agent, acs_proven: [], timestamp} These comments are written by implement-trd-beads when test tasks close with PASSED verdict.

2. Test Bead Status Scan Scan all test task beads for in-progress requirement verification. Run: br list --status=open --json If the call fails or returns non-JSON: print 'WARNING: Could not fetch open test beads — those tasks will appear in wrong status bucket.' Continue with empty list for that status. Run: br list --status=in_progress --json If the call fails or returns non-JSON: print 'WARNING: Could not fetch in_progress test beads — those tasks will appear in wrong status bucket.' Continue with empty list for that status. Run: br list --status=closed --json If the call fails or returns non-JSON: print 'WARNING: Could not fetch closed test beads — those tasks will appear in wrong status bucket.' Continue with empty list for that status. Filter for beads with title matching [trd:<TRD_SLUG>:task:*-TEST] pattern. For each test bead: extract task ID, read br native status (open/in_progress/closed). Read bead comments for req-satisfied: tokens. Parse each bead comment for 'req-satisfied:' substring within status:closed comment lines (the token is embedded as: 'status:closed qa: verdict:passed req-satisfied:<REQ_ID> ac-proven:'). Do not look for standalone req-satisfied: comment lines. Build TEST_TASKS map: TRD-NNN-TEST -> {status, req_satisfied, acs_proven}

Phase 3: Report Generation

1. Status Table Generation Generate requirement satisfaction table.

Cross-reference PRD_REQUIREMENTS (if available) with VERIFIED_REQS and TEST_TASKS:

For each REQ-NNN in PRD_REQUIREMENTS (or VERIFIED_REQS if no PRD): Status determination: SATISFIED: REQ-NNN in VERIFIED_REQS (test bead closed with PASSED) IN PROGRESS: corresponding -TEST bead is in_progress or in_review or in_qa PENDING: corresponding -TEST bead is open (not yet started) NOT PLANNED: no -TEST bead found for this requirement

If multiple test tasks satisfy the same REQ-NNN: status is SATISFIED only if ALL corresponding -TEST beads are closed-PASSED. If any are open or in_progress, status is IN PROGRESS. If any are closed but not PASSED, note the failure.

Print table: === REQUIREMENT SATISFACTION STATUS: <TRD_SLUG> ===

REQ-NNN	Description	Status	Test Task	ACs Proven
REQ-001	...	SATISFIED	TRD-001-TEST	AC-001-1, AC-001-2
REQ-002	...	IN PROGRESS	TRD-007-TEST	—
REQ-003	...	PENDING	TRD-003-TEST	—

Summary: SATISFIED: / IN PROGRESS: PENDING: NOT PLANNED:

Expected Output

Format: Requirement satisfaction table (console output)

Structure:

• Requirement Status Table: Per-requirement status (SATISFIED, IN PROGRESS, PENDING, NOT PLANNED) with test task and ACs proven
• Summary Counts: Aggregate counts by status category

Usage

/ensemble:requirement-status [trd-path-or-slug]