syndicate-status

/syndicate-status

You are inspecting the syndicate pipeline. Do not mutate state.

Step 1 — locate the repo

Resolve $SYNDICATE_REPO. If unset, fall back to $(pwd) and verify a pipeline/ directory exists there. If neither resolves, tell the user to export SYNDICATE_REPO=/path/to/syndicate in their shell rc and stop.

Step 2 — fetch the snapshot

Run exactly:

cd "${SYNDICATE_REPO:-$(pwd)}" && uv run python -m pipeline.cli status

The command emits a single JSON object on stdout:

{
  "ok": true,
  "result": {
    "db_path": "db/snapshot.db",
    "db_exists": true,
    "db_size_bytes": 12345678,
    "items_total": 1234,
    "items_last_24h": 42,
    "items_unsummarized": 7,
    "clusters_total": 89,
    "last_run_per_channel": {
      "gmail":   {"finished_at": "...", "ok": true,  "fetched": 12, ...},
      "rss":     {...},
      "twitter": {...}
    },
    "log_path_today": "logs/2026-05-15.txt",
    "log_tail": ["...", "..."],
    "disk_free_gb": 42.1,
    "ollama_reachable": true,
    "env_present": {"GMAIL_USER": true, "ANTHROPIC_API_KEY": false, ...},
    "git_branch": "main",
    "git_dirty": false
  },
  "log_path": null
}

Schema definition: pipeline/status.py StatusSnapshot.

Step 3 — render

Produce a compact, scannable summary:

• Header: syndicate · <items_total> items · +<items_last_24h> 24h · <clusters_total> clusters · <items_unsummarized> pending
• Per-channel table: one row each for gmail / rss / twitter from last_run_per_channel. Format: ✓ gmail · 12 saved · 2h ago or ✗ rss · failed: 502 (3h ago).
• Health: bullet any of these that are red:
  • ollama_reachable is false
  • disk_free_gb < 5
  • any required env var in env_present is false (GMAIL_USER, FEED_REPO_URL, OLLAMA_HOST for the default stack)
  • git_dirty is true (advisory only)
• Last log lines: include the last 5 entries of log_tail only if the most recent channel run failed, otherwise omit entirely.