epistex-dev/poteto-kit / templates/skills/verify-app
verify-app
UnverifiedRun the real {{APP_NAME}} the way users reach it, drive it, and capture evidence (screenshot, console errors, performance trace, heap snapshot). Use to reproduce a bug report, verify a fix or feature before claiming it works, or measure performance. Never report "should work" without evidence from a real run.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 18.3 KB |
| references/cli-contract.md | markdown | 9.8 KB |
| references/cloud-environment.md | markdown | 5.3 KB |
| references/evidence-export.md | markdown | 2.6 KB |
| references/features/README.md | markdown | 8.9 KB |
| references/features/_feature.template.md | markdown | 3.1 KB |
| references/features/cross-cutting-flows.md | markdown | 3.3 KB |
| references/launch-by-app-type.md | markdown | 2.7 KB |
| references/test-account.md | markdown | 3.4 KB |
| references/worktree-isolation.md | markdown | 5.2 KB |
references/test-account.md
View on GitHub<!-- HOW TO USE:
1. Keep this file only if the app has a login. Fill the {{PLACEHOLDERS}}; rename `account create` and the flags to match your CLI.
2. The behavior here is a contract for the control skill's CLI: build it once, test it (the overwrite refusal and the no-print rule especially), then let agents rely on it.
3. SKILL.md links here from "Prerequisites". Delete this comment block when done.
-->
# Test account: provisioning and sessions
Why: an automated run that logs in needs credentials, and credentials are where verification tooling leaks or damages things. One dedicated account, created by the tool, kept outside the repo and never shown, removes the temptation to reuse a real person's login or a password copied from an old script.
## Preferred: a CLI command that creates it
`{{CLI_NAME}} account create` registers one dedicated test account through the app's own sign-up flow, on a non-production environment:
- the email is unique and identifiable as a test account (a recognizable prefix plus a timestamp, on a domain you control or a reserved test domain), so it can be found and removed later;
- the password is strong and random, generated by the command (never chosen by the agent, never reused from elsewhere);
- credentials go to `{{CREDENTIALS_FILE}}`, a user config file outside the repo, directory mode 700 and file mode 600. The environment may override it at run time; the CLI reads the environment first;
- it refuses to overwrite an existing file unless `--force` is passed, and refuses when the environment already supplies credentials;
- it never prints the password, not even on failure, and redacts it from every log, report and artifact; write the file before registering, so a registered account cannot lose its password, and roll the file back if registration fails;
- it confirms with one login, and reports success or the HTTP status only.
Ask the developer before creating an account on any shared environment (staging used by others, a demo backend): it is a write on something that is not only theirs. On a local or disposable backend, no approval is needed.
## Around it
- `doctor` reports whether credentials are present (and from where: environment or file) and nothing else: no values, no length, at most a masked account name. It warns when the file is more permissive than mode 600.
- A run without credentials refuses (exit 3) and says to run `account create` or set the variables.
- Log the account in the adoption log (or the PR) by email only, with how to remove it: which identity store to delete it from and who has access. Never the password.
- If existing scripts in the repo hard-code credentials, tell the developer and do not copy, read out, reuse or "clean up" them yourself; the account belongs to someone and the leak is theirs to rotate.
## One session per account
Some backends allow a single session per account: a second login revokes the first. Two runs at once then log each other out and fail in confusing ways. Choose one policy, implement it, and have `doctor` say which one applies:
- **Serialize.** The CLI holds a run lock (a lock file with the owner's pid; a stale lock of a dead pid is replaced) and refuses a second run, exit 3. Simple, and right for one shared account.
- **One account per parallel worker.** Each worker has its own credentials entry. Needed only when runs must overlap.
Also keep humans off the test account while a run is going, and say so in the feature map's baseline preconditions.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |