Search

PythonReactSEOPDF
OpenClaw

auth-guard

Standardize API credential handling and startup auth checks to prevent "missing key" regressions across sessions. Use when an agent repeatedly loses auth state, gets intermittent 401/403 errors after restarts, relies on ad-hoc curl calls, or needs a reusable auth-first pattern for HEARTBEAT.md/AGENTS.md and helper scripts.

adainthelab0 source installs☆ 0 Repository stars

Installation

Follow the repository instructions for this skill and your agent. Review the files and dependencies before adding it to your workflow.

SkillHub CLI

skillhub install adainthelab/auth-guard

Inspect the source and compatibility requirements, then follow its installation steps.

Give your agent the source link and ask it to inspect instructions and dependencies before explaining the installation steps.

View source ↗Guides

Skill instructions

Auth Guard

Enforce a deterministic auth path: one credential source, one helper command path, one startup check, one fallback policy.

Quick Workflow

  1. Identify the target service endpoint and current failing flow.
  2. Define canonical credential source (env var first, credentials file second).
  3. Create/update a helper script in workspace (.pi/) that always injects auth.
  4. Add a startup/auth-check command that verifies credentials and endpoint access.
  5. Update HEARTBEAT.md or AGENTS.md to require helper usage (ban raw unauthenticated calls).
  6. Add explicit fallback behavior for unauthorized states.

Rules to Apply

  • Prefer ENV_VAR override, then ~/.config/<service>/credentials.json.
  • Never embed secrets in logs, memory notes, or chat responses.
  • Never call protected endpoints via raw curl if a helper exists.
  • Keep fallback behavior explicit and low-noise.
  • Store helper scripts in workspace/.pi/ for easy reuse.

Runtime Requirements

  • bash
  • curl
  • python3

Check once before using this skill:

command -v bash curl python3 >/dev/null

Safety Limits

  • Pass only trusted credential paths under ~/.config/<service>/... by default.
  • Do not point --cred-file at arbitrary workspace files or unrelated secret stores.
  • Keep probe URLs scoped to the target service auth endpoint.

Startup Auth Check Pattern

Run at session start (or before heartbeat loops):

bash skills/auth-guard/scripts/auth_check.sh \
  --service moltbook \
  --url 'https://www.moltbook.com/api/v1/feed?sort=new&limit=1' \
  --env-var MOLTBOOK_API_KEY \
  --cred-file "$HOME/.config/moltbook/credentials.json"

Expected outcomes:

  • AUTH_OK → proceed with normal authenticated helper flow.
  • AUTH_MISSING or AUTH_FAIL_* → use defined fallback path and record one concise note.

Reusable Snippets

Use drop-in policy snippets from:

  • references/snippets.md (HEARTBEAT + AGENTS + helper policy blocks)

References

  • `references/contrac...

Source information

Source
github
Repository
openclaw/skills
Path
skills/adainthelab/auth-guard
Version
—
Commit
a655a8ad58fdf9f17b9b4443d3d10a6cd3b1cb42
Classification source
rules
Status
active
Visibility
public

These signals come from indexed source analysis and may lag behind changes. Check the current source files.

Extracted content

SKILL.md
# Auth Guard

Enforce a deterministic auth path: one credential source, one helper command path, one startup check, one fallback policy.

## Quick Workflow

1. Identify the target service endpoint and current failing flow.
2. Define canonical credential source (env var first, credentials file second).
3. Create/update a helper script in workspace (`.pi/`) that always injects auth.
4. Add a startup/auth-check command that verifies credentials and endpoint access.
5. Update HEARTBEAT.md or AGENTS.md to require helper usage (ban raw unauthenticated calls).
6. Add explicit fallback behavior for unauthorized states.

## Rules to Apply

- Prefer `ENV_VAR` override, then `~/.config/<service>/credentials.json`.
- Never embed secrets in logs, memory notes, or chat responses.
- Never call protected endpoints via raw curl if a helper exists.
- Keep fallback behavior explicit and low-noise.
- Store helper scripts in `workspace/.pi/` for easy reuse.

## Runtime Requirements

- `bash`
- `curl`
- `python3`

Check once before using this skill:

```bash
command -v bash curl python3 >/dev/null
```

## Safety Limits

- Pass only trusted credential paths under `~/.config/<service>/...` by default.
- Do not point `--cred-file` at arbitrary workspace files or unrelated secret stores.
- Keep probe URLs scoped to the target service auth endpoint.

## Startup Auth Check Pattern

Run at session start (or before heartbeat loops):

```bash
bash skills/auth-guard/scripts/auth_check.sh \
  --service moltbook \
  --url 'https://www.moltbook.com/api/v1/feed?sort=new&limit=1' \
  --env-var MOLTBOOK_API_KEY \
  --cred-file "$HOME/.config/moltbook/credentials.json"
```

Expected outcomes:
- `AUTH_OK` → proceed with normal authenticated helper flow.
- `AUTH_MISSING` or `AUTH_FAIL_*` → use defined fallback path and record one concise note.

## Reusable Snippets

Use drop-in policy snippets from:
- `references/snippets.md` (HEARTBEAT + AGENTS + helper policy blocks)

## References

- `references/contrac...

Related skills

OpenClaw

bambu-studio-ai

heyixuan2

Bambu Lab 3D printer control and automation. Activate when user mentions: printer status, 3D printing, slice, analyze model, generate 3D, AMS filament, print monitor, Bambu Lab, or any 3D printing task. Full pi...

OpenClaw

humanpages

human-pages-ai

Search and hire real humans for tasks — photography, delivery, research, and more

OpenClaw

geo-optimization

capt-marbles

Generative Engine Optimization (GEO) for AI search visibility. Optimize content to appear in ChatGPT, Perplexity, Claude, and Google AI Overviews. Use when optimizing websites, pages, or content for LLM discove...

OpenClaw

openclaw-skills

zseven-w

Reusable skill templates for OpenClaw AI agents. Templates for API integration, data processing, web scraping, CLI tools, and file processing.