ReddRadar
Agent skill
uninstall-manifest-plugin
Remove Manifest from OpenClaw and reset to a direct model provider. Use when the user says "/uninstall-manifest-plugin", "uninstall manifest", "remove manifest", "reset openclaw", or wants to stop using Manifest routing and go back to direct provider access. Works for both cloud users (provider config only) and local users (plugin + provider config).
Install this agent skill to your Project
npx add-skill https://github.com/mnfst/manifest/tree/main/.claude/skills/uninstall-manifest-plugin
SKILL.md
Uninstall Manifest
Remove Manifest from OpenClaw -- both the model provider config and the plugin (if installed). Resets the default model to the best available provider.
Cloud users (no plugin): This removes models.providers.manifest from the config and resets the default model.
Local users (plugin installed): This also removes the plugin, auth profiles, and local data.
Workflow
1. Run the uninstall script
bash "${CLAUDE_SKILL_DIR}/scripts/uninstall_manifest.sh"
The script removes the provider config, plugin (if installed), auth profiles, and local data, then detects available providers and restarts the gateway. Use --dry-run to preview.
2. Set the default model
Read the script output for OUTPUT_PROVIDER and OUTPUT_MODEL lines. If a provider was detected, update the primary model:
jq --arg model "<OUTPUT_MODEL>" '
.agents.defaults.model.primary = $model
' ~/.openclaw/openclaw.json > /tmp/oc.json \
&& mv /tmp/oc.json ~/.openclaw/openclaw.json
Provider priority (first match wins):
| Provider | Env var | Default model |
|---|---|---|
| Anthropic | ANTHROPIC_API_KEY |
anthropic/claude-sonnet-4-6 |
| OpenAI | OPENAI_API_KEY |
openai/gpt-4o |
GOOGLE_API_KEY or GEMINI_API_KEY |
google/gemini-2.5-pro |
|
| DeepSeek | DEEPSEEK_API_KEY |
deepseek/deepseek-r1 |
| Groq | GROQ_API_KEY |
groq/llama-4-scout-17b |
If multiple exist, ask the user which to use as default.
3. Show status table
bash .claude/skills/manifest-status/scripts/manifest_status.sh
Output the table exactly as printed. No extra commentary.
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
mnfst/manifest
ensure-manifest-docs-consistency
Check consistency of Manifest documentation across all sources (GitHub, website, ClawhHub, npm, skills) against the actual codebase. Produces a dissonance report with tables. Use when the user says "/ensure-manifest-docs-consistency", "check manifest docs", "docs consistency", or wants to audit Manifest documentation for contradictions.
mnfst/manifest
manifest-status
Show current Manifest configuration as a diagnostic table. Use when the user says "/manifest-status", "manifest status", "show manifest config", "manifest settings", "is manifest installed", "check manifest", or wants to see the current Manifest routing setup. Outputs a table and nothing else.
mnfst/manifest
setup-manifest-plugin
Configure Manifest as a model provider in OpenClaw. Use when the user says "/setup-manifest-plugin", "setup manifest", "connect manifest", "point openclaw to localhost", or wants to add Manifest as a model provider. For cloud users, sets up models.providers.manifest directly (no plugin). For local dev, configures the plugin. Accepts a port number and optional mode.
mnfst/manifest
manifest
Model Router for OpenClaw. Save up to 70% by routing requests to the right model. Choose LLM fallback to avoid API rate limits, set thresholds and reduce token consumption.
openclaw/openclaw
openclaw-test-heap-leaks
Investigate `pnpm test` memory growth, Vitest worker OOMs, and suspicious RSS increases in OpenClaw using the `scripts/test-parallel.mjs` heap snapshot tooling. Use when Codex needs to reproduce test-lane memory growth, collect repeated `.heapsnapshot` files, compare snapshots from the same worker PID, distinguish transformed-module retention from real data leaks, and fix or reduce the impact by patching cleanup logic or isolating hotspot tests.
openclaw/openclaw
parallels-discord-roundtrip
Run the macOS Parallels smoke harness with Discord end-to-end roundtrip verification, including guest send, host verification, host reply, and guest readback.
Didn't find tool you were looking for?