Skip to content
Agentmart
Submit a tool

sre-core

Core SRE agent workflow, preflight checklist, and anti-patterns.

Skill by ndx-au

IndexedMIT
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/ndx-au-sre-core

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Core SRE agent workflow, preflight checklist, and anti-patterns. Use this skill for any work inside a Sentinel Runtime Environment (initialized project or this sre-init source). Covers the standard agent loop (match skill → preflight → mutate via API → validate), vault/storage discipline, and the most common mistakes that waste turns or introduce risk.

Use cases: backend, productivity, agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/sre-core

What's inside

Skill name
sre-core
Description
Core SRE agent workflow, preflight checklist, and anti-patterns. Use this skill for any work inside a Sentinel Runtime Environment (initialized project or this sre-init source). Covers the standard agent loop (match skill → preflight → mutate via API → validate), vault/storage discipline, and the most common mistakes that waste turns or introduce risk.
License
MIT
Instructions
37 lines in SKILL.md · view on GitHub

Before installing

  • Instructions call the network

Files (1)

  • SKILL.md

SKILL.md (first lines)

# SRE Core Agent Workflow & Guardrails

## Standard Agent Loop (always follow)
0. **Match task to an `sn-*` or `sre-*` skill** (or the kit-local README / kits/AGENTS.md). Load it explicitly before proceeding.
1. **Read `LOCAL.md`** (if present in the checkout — it is gitignored and contains machine paths).
2. **Preflight**: `GET /api/config` and confirm `vault_initialized: true` before any `POST /st`, secret writes, or storage/backup mutations.
3. **Mutate via the HTTP API only** (Bearer token from `.env` or config). Prefer `curl` (or the sentinel CLI that talks to the same API). No ad-hoc sidecar scripts that touch `.sentinel/` state directly.
4. After kit or flow changes: run `sentinel sre validate -p .` (and `sentinel sre test <kit>/` where relevant).
5. **Update twins / docs**: for every changed `.lua` also update its `*.lua.md`; for behavior changes visible to operators, update the matching page under `docs/`.

## API Discipline
- Always use the live harness API (`HARNESS_URL` + `SENTINEL_API_KEY` from `.env`).
- **Never** set the harness process env var `SENTINEL_API`.
- Cloud credentials and storage secrets live **only** in the vault (reached via `POST /st`, `POST /sc`). They are never in git, `.env`, or Lua source.
- `GET /hl`, `GET /api/config`, `GET /fl`, `GET /st`, `GET /bk/plans` are safe read entrypoints.

## Common Mistakes (avoid these)
1. Harness operations without first loading the matching `sn-*` skill.
2. `POST /st` (or any secret write) when `vault_initialized` is false.
3. Putting cloud credentials or `SENTINEL_VAULT_PASSPHRASE` into `.env` or any committed file.
4. Using the env var `SENTINEL_API` on the harness process (breaks config loading).
5. Loading an entire operator `.env` (with secrets) into a systemd unit for the harness.
6. Listing concrete flows or files inside `kit.yaml`.
7. Implementing long-running archive/backup jobs inside Lua flows instead of using harness `/bk` plans + `sn.backup.*` observation.
8. Committing `.env` or the `.sentinel/` directory.
9. Patching the private harness for pure SRE/content work (only do this for real missing API surface).

## Quick Preflight Snippet
```bash
export SENTINEL_API_KEY=$(grep '^SENTINEL_API_KEY=' .env | cut -d= -f2-)
export HARNESS_URL=${HARNESS_URL:-http://127.0.0.1:8080}
curl -s -H "Authorization: Bearer $SENTINEL_API_KEY" "$HARNESS_URL/api/config" | jq '.vault_initialized'
```

See also: root `AGENTS.md` (global layout, runtime table, key routes), `skills/sre-lua`, `skills/sre-triggers`, `skills/sre-config`, `sn-storage`, `sn-backups`, and any `kits/*/AGENTS.md` or kit README for the current directory.

From ndx-au/sre-init, MIT.

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Jun 16, 2026 (4 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from ndx-au

Cite this page

sre-core skill. Agent Mart. https://agentmart.dev/skills/ndx-au-sre-core (last updated 2026-10-24)