49296d538e
Some checks failed
Architecture Lint / Linter Tests (pull_request) Successful in 36s
PR Checklist / pr-checklist (pull_request) Failing after 6m56s
Smoke Test / smoke (pull_request) Failing after 34s
Validate Config / YAML Lint (pull_request) Failing after 21s
Validate Config / JSON Validate (pull_request) Successful in 9s
Validate Config / Python Syntax & Import Check (pull_request) Failing after 1m40s
Validate Config / Shell Script Lint (pull_request) Failing after 48s
Validate Config / Cron Syntax Check (pull_request) Successful in 18s
Validate Config / Deploy Script Dry Run (pull_request) Successful in 20s
Validate Config / Playbook Schema Validation (pull_request) Successful in 38s
Architecture Lint / Lint Repository (pull_request) Has been cancelled
Validate Config / Python Test Suite (pull_request) Has been cancelled
Scheduler that auto-starts batch pipelines when inference is available. Features: - Checks inference provider availability (local Ollama, RunPod, OpenRouter) - Priority ordering: playground > training > knowledge > adversary > genome - Dependency rules (e.g., knowledge_mine waits for training_factory) - Daily token budget (5M default, configurable) - Peak-hour pausing (8am-10pm = interactive mode, no pipelines) - State persistence via ~/.hermes/pipeline_state.json - One pipeline per cycle to avoid overload Usage: python3 pipeline/nightly_scheduler.py --status python3 pipeline/nightly_scheduler.py --check # dry-run python3 pipeline/nightly_scheduler.py # live Cron: */30 22-5 * * * pipeline/nightly_scheduler.py Closes #624