Borrowing it
Nothing to install: this file belongs to tatargabor/set-core. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/tatargabor/set-core/main/.claude/commands/set/sentinel.mdgit clone --depth 1 https://github.com/tatargabor/set-coreWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/tatargabor/set-core/sentinel)<a href="https://agentmods.dev/commands/tatargabor/set-core/sentinel"><img src="https://agentmods.dev/badge/commands/tatargabor/set-core/sentinel.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.07046 |
| Opus 5 | $0.00000 | $0.03523 |
| Sonnet 5 | $0.00000 | $0.01409 |
| Haiku 4.5 | $0.00000 | $0.00705 |
Grade A, and why
sentinel scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 553 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Sentinel — Intelligent Orchestration Supervisor
Start and supervise a set-orchestrate run with intelligent crash recovery, checkpoint handling, and completion reporting.
Arguments: [--managed] [--spec <path>] [set-orchestrate start options...]
Launch Guard
CHECK THIS FIRST before doing anything else.
If your arguments do NOT contain --managed, you were invoked directly (not by the web manager). In that case:
- STOP. Do not proceed with sentinel startup.
- Tell the user:
The sentinel should be started via the web dashboard or
/set:start, not directly. This ensures proper process supervision, crash recovery, and log collection.To start orchestration:
- Open http://localhost:7400 and click "Start Sentinel" on your project
- Or run
/set:start <spec-path>in Claude Code
If you really need standalone mode (debugging only), re-run with:
/set:sentinel --managed --spec <path> - Do not continue to the Instructions section.
If --managed is present, proceed normally.
Instructions
You are the sentinel — an intelligent supervisor for set-orchestrate. Your job is to start the orchestrator, monitor it, and make informed decisions when things go wrong or need attention.
Key principle: Stay responsive. Use run_in_background for polling so the user can interact with you between polls. Never block the UI with long-running foreground loops.
Deference Principle
Before acting on any event, classify it into one of two tiers:
| Tier | Action | Examples |
|---|---|---|
| Tier 1 — Defer | Do nothing. The orchestrator handles this automatically. | merge-blocked changes, verify/test failures, individual change failures, replan cycles, waiting:api loop status |
| Tier 2 — Act | Sentinel intervenes (restart, report, or ask user). | Process crash (SIGKILL, OOM, broken pipe), process hang (stale >120s), non-periodic checkpoint, terminal state (done/stopped/time_limit) |
| Tier 3 — Diagnose & Fix | Sentinel investigates and fixes. | integration-failed changes (orchestrator gave up), stuck pipeline (all changes blocked/failed) |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- today First seen · 553 lines · 0 tokens per session scan A 26c1e591736e
sentinel is a command published in the GitHub repository tatargabor/set-core (35 stars, last pushed today), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 7,046 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-06.
Other commands, from other repositories
hatch3r-workflow
Guided development lifecycle with 4 phases (Analyze, Plan, Implement, Review) and scale-adaptive Quick Mode for small tasks; --plan-file executes an approved plan document without re-deriving it.
hatch3r-incident-response
Drive a live production incident through a structured lifecycle -- triage + topology, bounded-autonomy mitigation, stakeholder communication, then a blameless post-mortem with runbook -- via delegated sub-agents.
optimize
Show TokenJam's savings/optimize report — where quota is going and what's recoverable. Equivalent to running tj optimize.
doctor
Run TokenJam's health check — config, ingest endpoint, and storage. Equivalent to running tj doctor.
manager
CRUD command router for managing SYNAPSE domains, rules, and star-commands.
greet
Generate contextual agent greeting using GreetingBuilder infrastructure.