Repository navigation
Releases
Releases · grobomo/ccc-manager
v1.24.0 — Worker Task Claims & Env Var Config
What's New
Worker Task Claims (Multi-Worker Coordination)
File-based claim/release/isClaimed with automatic expiry. When multiple workers share a state directory, each worker claims tasks before processing. Other workers skip claimed tasks. Claims auto-expire (default 10min) so crashed workers don't permanently lock tasks.
workerId: worker-1
claimTimeout: 600000Config Env Var Interpolation
${VAR} and ${VAR:-default} syntax in YAML values. Secrets from environment instead of plaintext in config files — the standard K8s/Docker pattern.
notifiers:
teams:
url: ${TEAMS_WEBHOOK_URL}
secret: ${HMAC_SECRET:-dev-secret}Stats
- 22 test suites, 607 tests
- 20 built-in components
- 73 PRs merged
v1.19.0 — Docs & Polish
Changes
- README: component table updated from 17→20 (added parallelDispatch, aggregateResults)
- README: test counts updated from 476→581, suite counts 19→21
- Helm version reference updated to v1.18.0
Stats
- 20 components, 21 test suites, 581 tests
- 70 PRs merged
v1.18.0 — Test Coverage & Logger Fix
- Direct unit tests for exec-helper.js (18 tests)
- Logger JSON mode: filter undefined values before spread
- 21 suites, 581 tests total
v1.17.0 — Code Review DRY Pass
- DRY: shared execCommand helper for ProcessMonitor + CronMonitor
- Logger: filter undefined values from text output
- SQSInput.listen(): try/catch prevents silent poll loop death
- Full code review of all 26 source files — no other issues found
v1.16.0 — Security Hardening
- Security: sanitize bridge writeResult requestId to prevent path traversal
- Component table updated to 20 components
- 20 suites, 556+ tests
See CHANGELOG.md for details.
v1.15.0 — Sharding, SQS, Priority Dispatch
Highlights
- Task sharding with dependency graph and parallel execution
- Sharder module: cartesian/round-robin/chunk strategies
- Priority-aware queue (critical/high/normal/low)
- SQS input + SQS dispatcher for distributed AWS fleet ops
- Event-driven bridge with fs.watch
- Standalone dispatcher-api.py for ep-incident-response
- 20 components, 20+ suites, 525+ tests
EP Incident Response Unblocked
All 5 cross-project blockers (T148-T152) resolved.
See CHANGELOG.md for full details.
v1.14.0 — Helm Chart
Highlights
- Helm chart for one-command K8s deployment
- RONE values overlay for hackathon-teams-poller
- Chart test suite (49 tests)
- 17 suites, 444 tests total
See CHANGELOG.md for full details.
v1.13.0 — DRY MultiManager, README, Polish
Changes
- DRY:
MultiManager.start()delegates toManager.start()— removed 22-line duplication - README: Multi-instance CLI docs,
MultiManagerAPI, updated test counts (16 suites, 395 tests) - Rules migration: Archived
.claude/rules/— content moved to hook modules + CLAUDE.md
Stats
- 17 components, 16 test suites, 395 tests
- 62 PRs merged
v1.12.0 — Multi-Instance Support
Multi-Instance Support
Run multiple manager instances from a single process, each with isolated state and labeled metrics.
New Features
- Per-instance state isolation — each Manager uses
state/<name>/directory - Instance-labeled Prometheus metrics —
{instance="name"}labels on all metrics MultiManagerclass — shared health endpoint aggregating all instances- Multi-config CLI —
ccc-manager a.yaml b.yaml --health-port 8080 --health-portflag — configure shared health endpoint port--statusenhancement — shows per-instance state subdirs
Stats
- 16 test suites, 395 tests (36 new)
- 17 built-in components
- Zero npm dependencies
v1.11.0
What's New
- GitHub Actions test CI (Node 18/20/22 matrix)
- README.md with quickstart, architecture, component table
- Grafana dashboard included in Docker image
- GitHub issues input added to rone-teams-poller config
- Fix secret-scan self-match (PRIVATE KEY grep matched its own workflow file)
- Cleaned up 14+ stale remote branches
Stats
- 17 built-in components
- 15 test suites, 359 tests
- Zero npm dependencies