Skip to content

Business operating system for Claude Code — 57 skills, 21 agents, smart daemon. Unified inbox (WhatsApp/Email/Slack/Telegram), autonomous PR merge, full-AWS monitoring, revenue (Stripe+RevenueCat), e-commerce (Shopify), marketing (Klaviyo/Meta/GA4), voice (Bland/ElevenLabs), APM (Datadog/NewRelic/OTEL), YOLO mode.

docs(09-03): complete /ops:monitor + /ops:integrate plan summary

auroracapital3288d6a98801
diff --git a/.planning/phases/phase-09/09-03-SUMMARY.md b/.planning/phases/phase-09/09-03-SUMMARY.md
new file mode 100644
index 00000000..7c167c91
--- /dev/null
+++ b/.planning/phases/phase-09/09-03-SUMMARY.md
@@ -0,0 +1,106 @@
+---
+phase: "09"
+plan: "03"
+subsystem: "ops-monitor, ops-integrate, monitor-agent"
+tags: [monitoring, apm, datadog, newrelic, otel, saas-integration, partner-registry]
+dependency_graph:
+ requires: []
+ provides: [ops-monitor skill, ops-integrate skill, monitor-agent, monitoring userConfig keys]
+ affects: [ops router, plugin.json]
+tech_stack:
+ added: [Datadog API v1/v2, New Relic GraphQL API, OpenTelemetry healthz]
+ patterns: [agent-spawn-from-skill, partner-registry-jq, atomic-tmpfile-swap, haiku-polling-agent]
+key_files:
+ created:
+ - claude-ops/agents/monitor-agent.md
+ - claude-ops/skills/ops-monitor/SKILL.md
+ - claude-ops/skills/ops-integrate/SKILL.md
+ modified:
+ - claude-ops/skills/ops/SKILL.md
+ - claude-ops/.claude-plugin/plugin.json
+decisions:
+ - "monitor-agent uses claude-haiku-4-5 per D-01 decision — lightweight, frequent polling, no write access"
+ - "partner_registry stored in preferences.json (local, gitignored) not a separate file"
+ - "ops router adds monitor row only — settings/integrate rows already present from Plan 02"
+metrics:
+ duration: "~15 min"
+ completed: "2026-04-14"
+ tasks_completed: 2
+ tasks_total: 2
+ files_created: 3
+ files_modified: 2
+---
+
+# Phase 09 Plan 03: /ops:monitor + /ops:integrate Skills Summary
+
+**One-liner:** Haiku-4-5 APM probe agent + unified Datadog/New Relic/OTEL monitoring skill + generic SaaS API onboarding with local partner registry.
+
+## Tasks Completed
+
+| Task | Name | Commit | Files |
+|------|------|--------|-------|
+| 1 | Create monitor-agent + ops-monitor/SKILL.md + plugin.json userConfig | 70cf25d | agents/monitor-agent.md, skills/ops-monitor/SKILL.md, .claude-plugin/plugin.json |
+| 2 | Create ops-integrate/SKILL.md + update ops router | 70cf25d | skills/ops-integrate/SKILL.md, skills/ops/SKILL.md |
+
+## What Was Built
+
+### monitor-agent.md
+- Model: `claude-haiku-4-5`, `effort: low`, `maxTurns: 15`
+- `disallowedTools: [Write, Edit, Agent]` — fully read-only probe agent
+- Preflight reads credentials from `preferences.json` via `jq`
+- Datadog: polls `/api/v1/monitor` for Alert/Warn state + `/api/v2/apm/traces` for top error traces
+- New Relic: GraphQL query for `CRITICAL` severity entities
+- OTEL: `GET $OTEL_ENDPOINT/healthz` for collector health
+- Outputs structured JSON with `summary.severity` (critical/warning/healthy) and `backends_configured`
+- Keys passed as `-H` headers only — never in URLs or terminal output (T-09-03-01 mitigation)
+
+### ops-monitor/SKILL.md
+- `--setup`: AskUserQuestion (≤4 options) → collect credentials → atomic tmpfile write → smoke test
+- Default: spawns `monitor-agent` via Agent tool, displays formatted dashboard with ✅/⚠️/🔴/⬜ icons
+- `--watch`: 60-second poll loop, diff-based output showing 🆕 new / ✅ resolved
+- `--backend <name>`: filter to single backend
+- CLI/API reference table for all three backends
+
+### ops-integrate/SKILL.md
+- `--list`: renders partner registry as table
+- 5-step onboarding: discover → confirm (user-gated) → credential → health check → registry write
+- jq `--arg` parameterized writes prevent JSON injection (T-09-03-02 mitigation)
+- User confirms discovered URL before any credential collection or write (T-09-03-03 mitigation)
+- Atomic tmpfile swap for all preference writes
+- `partner_registry` stored in `preferences.json` (local only, gitignored)
+
+### ops/SKILL.md router
+- Added row: `monitor, apm, alerts, datadog, newrelic, otel → /ops:ops-monitor $ARGUMENTS`
+- `settings` and `integrate` rows were already present from Plan 02 — not duplicated
+
+### plugin.json
+- 5 new `userConfig` keys: `datadog_api_key` (sensitive), `datadog_app_key` (sensitive), `newrelic_api_key` (sensitive), `newrelic_account_id`, `otel_endpoint`
+- JSON validated: `jq .` passes
+
+## Deviations from Plan
+
+### Auto-fixed Issues
+
+**1. [Rule 1 - Deviation avoidance] ops router already had settings/integrate rows**
+- **Found during:** Task 2
+- **Issue:** Plan said to add all 3 rows but Plan 02 had already added settings/integrate
+- **Fix:** Added only the monitor row — checked with Grep before editing per plan instructions
+- **Files modified:** claude-ops/skills/ops/SKILL.md
+
+## Known Stubs
+
+None — all data flows wired. Agent spawns `monitor-agent` which reads live credentials and polls real APIs. Partner registry reads/writes from/to `preferences.json` at runtime.
+
+## Threat Flags
+
+None beyond what is documented in the plan's threat model (T-09-03-01 through T-09-03-04, all mitigated in implementation).
+
+## Self-Check: PASSED
+
+- [x] `claude-ops/agents/monitor-agent.md` — exists, has `haiku-4-5`, has `disallowedTools`
+- [x] `claude-ops/skills/ops-monitor/SKILL.md` — exists, has --setup, --watch, default flows
+- [x] `claude-ops/skills/ops-integrate/SKILL.md` — exists, has `partner_registry`
+- [x] `claude-ops/skills/ops/SKILL.md` — has `ops-monitor`, `ops-settings`, `ops-integrate` routes
+- [x] `claude-ops/.claude-plugin/plugin.json` — valid JSON, has `datadog_api_key`
+- [x] lint: 200 passed, 0 failed
+- [x] commit: 70cf25d