Install
$ agentstack add skill-reopt-ai-reopt-skills-reopt-cli ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
reopt CLI
> This is NOT the reopt CLI you know. Run reopt --help / reopt --help for the live command tree, and read node_modules/@reopt-ai/cli/README.md for narrative guides (the CLI ships no dist/docs/).
When to apply
- Any
reoptCLI command. - Confirming login state, writing CI/CD steps around
reopt, scripting Brandapp OAuth credentials. - Load before
reopt-brandappandreopt-eav.
Step 1 — Pin agent rules into AGENTS.md / CLAUDE.md
Source: the CLI's own agent-rules file once it ships one (@reopt-ai/cli does not, as of 0.3.1). Fallback: agent-rules.md bundled with this skill. Wrap content between:
…content from source…
Idempotent: replace only between markers. This same block is shared by reopt-brandapp and reopt-eav — those skills will skip the step if the block is already present.
Step 2 — Auth (consumer-side credentials; CLI docs cover usage)
The CLI has three credential systems:
- User session —
reopt login/reopt status/reopt logout. Tokens live in~/.reopt/auth.json. Session auto-refreshes inside the Better Auth extension window; do not over-engineer retry loops in CI. - Brandapp OAuth credentials —
BRANDAPP_CLIENT_ID/BRANDAPP_CLIENT_SECRETenv vars, orreopt brandapp linkinteractively. Stored in~/.reopt/credentials.json+.reopt.json(project root). - Service token (CI/CD) —
reopt token mint --ttl 15m --scope eav:migrate,eav:readprints a short-lived, scoped JWT for headless pipelines; capture withexport BRANDAPP_CLIENT_TOKEN=$(reopt token mint --quiet). Pass to the SDK as a Bearertoken— never ship aclientSecretwhere a scoped token works.
Run reopt status (or reopt status --ping) before any mutating operation. If auth: not logged in, run reopt login first.
Step 3 — Route to module docs / --help
Prefer --help (live source of truth); the CLI ships no dist/docs/, so narrative detail lives in node_modules/@reopt-ai/cli/README.md:
| Task signal | Read | |---|---| | Any command / flag | reopt --help (live) | | Auth commands and session model | README.md § Authentication | | Service-token issuance (CI/CD) | reopt token mint --help | | Brandapp ops (link, doctor, init, dev, env, …) | reopt brandapp --help + see reopt-brandapp skill | | EAV ops (status, sync, pull, diff, plan, migrate, history, verify) | reopt brandapp eav --help + see reopt-eav skill | | Schema-as-Code, MCP, completion, config | README.md §§ Schema-as-Code, Shell completion, Preferences | | Global flags, output formats, pagination | README.md § Output and global flags | | Exit codes | README.md § Exit codes |
Quick global-flag reminders (subset; full list in --help):
--format json|table|csv|yaml— JSON is the agent-friendly default.--page-all+--page-limit+--page-delay— paginated iteration; JSON output becomes NDJSON.--no-interactive— required for unattended scripts (fail instead of prompt).--dry-run— preview only (EAV sync, brandapp link/unlink).
Exit code summary: 0 ok, 1 API/network, 2 auth, 3 validation, 4 config, 5 internal. EAV migrate/verify add 6 drift-detected, 7 destructive-blocked, 8 checksum-mismatch, 9 checksum-conflict, 10 lock-held.
Safety
- Never hardcode credentials.
- Never print credential values.
- Never commit files under
~/.reopt/. - Inject secrets through a secret manager in CI.
- Use
--dry-runbefore any mutating EAV operation, and especially before--delete-orphans. - Pass arguments as arrays when invoking the CLI programmatically.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: reopt-ai
- Source: reopt-ai/reopt-skills
- License: MIT
- Homepage: https://skills.sh/reopt-ai/reopt-skills
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.