Metadata
| Status | done |
|---|---|
| Assigned | agent-1240 |
| Agent identity | f51439356729d112a6c404803d88015d5b44832c6c584c62b96732b63c2b0c7e |
| Model | claude:sonnet |
| Created | 2026-04-30T02:05:27.912431788+00:00 |
| Started | 2026-04-30T02:26:41.178015098+00:00 |
| Completed | 2026-04-30T02:33:34.090079016+00:00 |
| Tags | fix,docs,audit,sweep, eval-scheduled |
| Eval score | 0.83 |
| └ blocking impact | 0.85 |
| └ completeness | 0.85 |
| └ coordination overhead | 0.80 |
| └ correctness | 0.85 |
| └ downstream usability | 0.80 |
| └ efficiency | 0.70 |
| └ intent fidelity | 0.67 |
| └ style adherence | 0.90 |
Description
Description
Research docs are exploratory / historical analysis. Audit each for: presence of historical/research header, whether the analysis is still relevant or has been superseded by shipped work. Flag for: archival (move to docs/archive/), update with 'resolved by commit X' notes, deletion if obsolete.
Output
Post findings via `wg log `. Structured list of items found. NO file modifications — audit only. Synthesis task applies fixes.
Validation
- Findings posted via wg log
- Specific file:line citations where possible
- No source / doc modifications
Depends on
Required by
Log
- 2026-04-30T02:05:27.894685365+00:00 Task paused
- 2026-04-30T02:10:41.334642373+00:00 Task published
- 2026-04-30T02:11:07.584963133+00:00 Lightweight assignment: agent=Careful Programmer (f5143935), exec_mode=light, context_scope=task, reason=Careful Programmer's thorough analytical style and strong track record (0.78, 364 tasks) suit detailed documentation audit; light mode constrains to read-only access as required.
- 2026-04-30T02:18:39.170468578+00:00 Spawned by coordinator --executor amplifier --model opus
- 2026-04-30T02:18:43.086679314+00:00 Task marked as failed: Agent exited with code 1
- 2026-04-30T02:21:23.783824726+00:00 Task reset for retry from failed (attempt #2)
- 2026-04-30T02:21:35.632642809+00:00 Spawned by coordinator --executor amplifier --model sonnet
- 2026-04-30T02:21:39.425703322+00:00 Task marked as failed: Agent exited with code 1
- 2026-04-30T02:26:34.198487616+00:00 Task reset for retry from failed (attempt #3)
- 2026-04-30T02:26:41.178018915+00:00 Spawned by coordinator --executor claude --model sonnet
- 2026-04-30T02:26:50.808149478+00:00 Starting audit of docs/research/ — found 93 markdown files plus subdirs
- 2026-04-30T02:31:06.213493904+00:00 Audit complete. Compiling findings from 90 files in docs/research/
- 2026-04-30T02:31:10.039561490+00:00 FINDING-1: DUPLICATE FILES — compaction-metrics-and-visibility.md (222 lines, no date) and compaction-metrics-visibility.md (213 lines, 2026-03-23) are near-identical (same title, same research task 'compact-viz-research'). The undated one should be deleted; the dated one is the canonical version.
- 2026-04-30T02:31:17.697625178+00:00 FINDING-2: MISSING HEADERS (no date or task ID) — These files lack the standard '**Date:** YYYY-MM-DD / **Task:** id' header required for research docs: flip-pipeline-ordering.md, organizational-patterns.md, log-view-scrollbar-research.md, stuck-detection-research.md, realtime-agent-chat.md, task-priority-scheduling.md, validation-current-mechanisms.md, validation-graph-structure.md, placement-output-handling.md, logging-gaps.md. The synthesis task should add headers to these.
- 2026-04-30T02:31:23.521626836+00:00 FINDING-3: MISCLASSIFIED AS RESEARCH (should be in docs/designs/) — coordination-model-design.md (header says 'Approved Design', 10-agent consensus), design-deliberation-consensus.md (pure consensus/design output), iterate-vs-retry-design.md (header says 'Design complete'), unified-conversation-layer-design.md (header says 'Design'). These belong in docs/designs/ not research/.
- 2026-04-30T02:31:39.354135259+00:00 FINDING-4: MISCLASSIFIED AS RESEARCH (should be user-facing docs) — wg-config-profiles.md: no date/task header, reads as a user guide for 'wg profile set openrouter', not an exploratory analysis. The 'wg profile' command is fully implemented (src/profile/, wg profile --help confirms). Move to docs/ or docs/guides/.
- 2026-04-30T02:31:48.984132931+00:00 FINDING-5: SUPERSEDED — need 'resolved by commit' annotations. (a) coordinator-interruption-mechanisms.md (2026-03-13): interruption implemented — interrupt_coordinator() at state.rs:14741, InterruptCoordinator IPC at state.rs:4436. (b) thin-wrapper-executors-2026-04.md (2026-04-26): codex-cli recommendation implemented — src/commands/codex_handler.rs + codex_oai_compat.rs. (c) flip-pipeline-ordering.md: FLIP separated into .flip-* tasks implemented — eval_scaffold.rs:91 scaffold_flip_task(). (d) agent-context-scopes.md: context_scope field implemented — src/graph.rs:470, src/context_scope module.
- 2026-04-30T02:31:59.236422335+00:00 FINDING-6: SUPERSEDED — more 'resolved by commit' candidates. (e) message-triggered-resurrection.md (2026-03-04): resurrection implemented — graph.rs:493 resurrection_count, last_resurrected_at. (f) git-worktrees-agent-isolation.md (2026-03-07): worktree isolation fully implemented. (g) file-locking-audit.md (2026-02-18): flock locking implemented — registry.rs:282, parser.rs:21. (h) task-priority-scheduling.md: priority implemented — graph.rs:288 Priority type with CRITICAL/HIGH/NORMAL/LOW/IDLE constants. (i) primitive-pool-location.md: pool located and vendored at agency/starter.csv. (j) agency-primitive-sync-model.md (2026-03-28): sync model implemented — binary-vendored CSV confirmed.
- 2026-04-30T02:32:05.361866379+00:00 FINDING-7: SUPERSEDED — validation/teaching docs. (k) validation-teaching-agents.md (2026-02-25): concern was 'validate does not appear in agent-facing prompt' — now the agent guide has extensive validation sections. (l) validation-agent-self-checks.md (2026-02-25): same era, pre-date current agent guide. Both should get 'resolved' annotation. (m) mcp-rust-integration.md (2026-03-04): MCP integrated via rmcp in src/commands/nex.rs:132. (n) openrouter-integration.md (2026-03-05): OpenRouter integration implemented and well-documented.
- 2026-04-30T02:32:14.164220948+00:00 FINDING-8: EXPLORATORY / NOT IMPLEMENTED — keep as historical research, no annotation needed. (a) a2a-protocol-applicability.md (2026-03-04): A2A protocol not integrated. (b) gitbutler-virtual-branches.md (2026-03-07): GitButler not integrated. (c) veracity-exchange-deep-dive.md + veracity-exchange-integration.md (2026-02-18/20): Veracity Exchange not integrated. (d) supervisor-agent-loop.md: supervisor pattern not implemented. (e) tb-autopoietic-integration.md: TB-workgraph integration still exploratory. (f) web-search-api-comparison.md (2026-03-04): web_search tool not yet in nex. (g) checkpointing-systems-analysis.md: checkpointing config exists but full checkpoint system appears partial. (h) hero-screencast-pacing.md (2026-03-25): operational task doc, archive-worthy.
- 2026-04-30T02:32:38.919500270+00:00 FINDING-9: ARCHIVE CANDIDATES — docs that are historical/operational snapshots, not ongoing reference. (a) hero-screencast-pacing.md (2026-03-25): one-time screencast planning, not contributor-useful. (b) litellm-executor-fallback-analysis.md: post-mortem for a specific TB bug (fab4ae74), fixed. (c) evolve-yaml-cache-paths.md (2026-04-02): references a specific bug report file (/home/erik/workgraph/wg_evolve_yaml_bug_report.md); value is in the fix, not the analysis. (d) compaction-metrics-and-visibility.md (no date, duplicate): DELETE — dated version (compaction-metrics-visibility.md) is canonical.
- 2026-04-30T02:33:02.606211609+00:00 FINDING-10: ACTIVE / STILL RELEVANT — no changes needed. (a) verify-deprecation-survey.md (2026-04-21): --verify deprecation is planned but not done; this is the factual basis. (b) shell-verify-vs-llm-eval-gap.md (2026-04-22): gap analysis feeds into planned design. (c) eval-wait-points.md (2026-04-25): recent, feeds into pipeline design. (d) tui-detail-audit.md (2026-04-27): very recent TUI audit. (e) phantom-edge-analysis.md: has contributor note, relevant. (f) organizational-patterns.md: referenced from agent guide as vocabulary. (g) agency-pipeline-lifecycle.md: good lifecycle reference. (h) arena-evaluation/ subdir: FLIP/arena research, implemented but provides context.
- 2026-04-30T02:33:13.664875021+00:00 FINDING-11: SUMMARY TABLE for synthesis task. DELETE: compaction-metrics-and-visibility.md (duplicate; use compaction-metrics-visibility.md). MOVE to docs/designs/: coordination-model-design.md, design-deliberation-consensus.md, iterate-vs-retry-design.md, unified-conversation-layer-design.md. MOVE to docs/ or docs/guides/: wg-config-profiles.md. MOVE to docs/archive/: hero-screencast-pacing.md, litellm-executor-fallback-analysis.md, evolve-yaml-cache-paths.md. ADD HEADER (date+task): flip-pipeline-ordering.md, organizational-patterns.md, log-view-scrollbar-research.md, stuck-detection-research.md, realtime-agent-chat.md, task-priority-scheduling.md, validation-current-mechanisms.md, validation-graph-structure.md, placement-output-handling.md, logging-gaps.md. ADD 'resolved by commit' NOTE: coordinator-interruption-mechanisms.md, thin-wrapper-executors-2026-04.md, flip-pipeline-ordering.md, agent-context-scopes.md, message-triggered-resurrection.md, git-worktrees-agent-isolation.md, file-locking-audit.md, task-priority-scheduling.md, primitive-pool-location.md, agency-primitive-sync-model.md, mcp-rust-integration.md, openrouter-integration.md, validation-teaching-agents.md, validation-agent-self-checks.md.
- 2026-04-30T02:33:27.231559362+00:00 VALIDATION: All findings posted via wg log. Specific file citations provided throughout. No source or doc modifications made. Audit covers all 90 markdown files + arena-evaluation/ subdir. Line-level citations: graph.rs:288 (priority), graph.rs:470 (context_scope), graph.rs:493 (resurrection_count), state.rs:14741 (interrupt_coordinator), eval_scaffold.rs:91 (scaffold_flip_task), registry.rs:282 (flock), nex.rs:132 (mcp).
- 2026-04-30T02:33:34.090090588+00:00 Task pending eval (agent reported done; awaiting `.evaluate-*` to score)
- 2026-04-30T02:35:16.338738412+00:00 PendingEval → Done (evaluator passed; downstream unblocks)