Internal infrastructure review and architecture decisions
The CSA pipeline transforms unstructured project inputs into deployed HTML artifacts with structured decision capture through a 12-phase rigid skill chain.
Total Engagements
Files Produced
Workspace Size
Ext Dependencies
Fig 1: Pipeline phase flow with dependency classification
| Dependency | Endpoint | Critical? | SLA |
|---|---|---|---|
| FTP Server | ftp://151.106.96.76 | YES | ~99.5% |
| Mermaid API | chart.chem.dev/api/v1 | YES | ~99.9% |
| Slack Notify | io.noboxai.com/notify.php | NO | ~99.5% |
| Playwright | Local | NO | Local |
The pipeline has 4 external dependencies. Currently, any single failure blocks the entire run. Implement graceful degradation with critical/non-critical classification.
Fig 2: Graceful degradation flow for partial failures
Dashboard is manually maintained. At 18 projects and growing, a hybrid manifest-driven approach would reduce maintenance while preserving control.
Fig 3: Manifest-driven dashboard generation
Phase 10.1 was just added. _handoff.json should be required for all new engagements, with optional backfill for existing 18.
Fig 4: _handoff.json rollout timeline
These decisions shape the next iteration of the CSA pipeline system. Each has a recommendation based on current operational patterns.
| # | Decision | Options | Recommended |
|---|---|---|---|
| D1 | Partial failure handling | Halt-all / Graceful degrade / Retry queue | Graceful degrade |
| D2 | _handoff.json requirement | All / New only / Optional | New only |
| D3 | Flywheel version numbering | Sequential / Date-based / Hybrid | Sequential |
| D4 | Quality thresholds | QA only / Score+age / Score+resolved% | Score + resolved % |
| D5 | Dashboard generation | Full auto / Manual / Hybrid manifest | Hybrid manifest |
Fig 5: Quality threshold decision flow for flywheel vs. fresh run
| Current Item | Unlocks | Gate |
|---|---|---|
| D1 (Partial failure) | LangGraph reimplementation error model | Failure handling pattern defined |
| D2 (Handoff manifest) | GitReport CSA-ADW conversion detection | _handoff.json schema finalized |
| D5 (Dashboard gen) | Response analytics dashboard | Manifest includes response metadata |
| 1.3 (Manifest rollout) | Multi-operator support | Manifest = portable engagement state |
Code-based CSA pipeline with typed state, interrupt gates, and parallel fan-out. Already spec'd at io.noboxai.com/csa-lg-architecture/.
Pipeline is single-operator. No auth, no role separation. Activates when a second operator needs independent CSA pipeline access.
Cross-engagement response analytics. Activates when 5+ engagements complete flywheel cycles with captured responses.
| Risk | Impact | Likelihood | Mitigation |
|---|---|---|---|
| FTP host downtime | Deploy blocked | Low | Stage locally, retry after 30min |
| Mermaid API breaking change | Diagrams fail | Low | Pin server version, test before upgrade |
| SKILL.md drift from behavior | Pipeline errors | Medium | Monthly full-pipeline integration test |
| Dashboard row count | UX degradation | Medium | Auto-gen with pagination at 30 projects |
| _handoff.json schema change | Inconsistency | Low | Version schema, migrate on read |