AIMA Doctor

Diagnose and repair OpenClaw when your agent is stuck, confused, or failing. One-command rescue for process health, config breakage, and device registration.

Security Scan

The skill's requested actions and install behavior are coherent with a local repair tool, but it downloads and executes a packaged runtime from an external release which has some minor inconsistencies you should verify before installing.

Purpose & Capability

The name/description (diagnose and repair OpenClaw) matches the runtime actions: the skill checks for a local helper and, if absent, downloads and runs a packaged runtime to perform repairs. Asking to place a helper under ~/.openclaw/tools/aima-doctor/ and to run it is proportionate to the stated purpose.

Instruction Scope

Instructions are narrowly scoped to checking/placing/executing the helper runtime and relaying user symptoms. The SKILL.md explicitly warns not to solicit remote shell one-liners and to keep tokens out of chat. However, the runtime is executed with a symptom string provided by the agent — the helper could access system state, so you should inspect the runtime before running if you have concerns.

Install Mechanism

This is instruction-only and the install step directs automatic download from a GitHub releases URL with an expected SHA256 checksum (good practice). A minor inconsistency: README references a different host (aimaservice.ai/doctor/runtime.zip) while SKILL.md points to the GitHub releases URL. Downloads from GitHub releases are common; extracting and executing an archive into the user's home is higher-risk but coherent for a repair runtime.

Credentials

The skill requests no environment variables or credentials and does not require unrelated secrets. The SKILL.md also admonishes keeping sensitive tokens out of chat. This is appropriate for a local repair tool.

Persistence & Privilege

The skill writes the runtime under ~/.openclaw/tools/aima-doctor/ and executes it, which is normal for a helper. always is false and agent autonomous invocation remains allowed (platform default). There is no indication the skill modifies other skills or system-wide settings, but the downloaded runtime will run arbitrary code locally — that persistence and privilege should be acknowledged.

Assessment

This skill appears to do what it says: download a packaged helper and run it to diagnose and repair OpenClaw. Before installing, verify the release/source: 1) Confirm the GitHub repository and homepage URLs are legitimate and match the vendor you trust. 2) If the installer runs automatically, prefer the SHA256 verification step and manually verify the checksum yourself. 3) Note README references a different download host — prefer the canonical release location you trust and avoid ambiguous mirrors. 4) Inspect the extracted runtime (run.sh / run.ps1) before executing, or run it in a sandbox/VM if you're unsure. 5) Never paste secrets into chat unless explicitly asked and you understand why; follow the skill's guideline to keep tokens out of normal chat. If you want greater assurance, request the runtime source code or a signed release from the publisher before proceeding.

Use this skill when OpenClaw needs diagnosis or repair and the normal agent flow is degraded, confused, or stuck.

Runtime contract

• The helper runtime lives at ~/.openclaw/tools/aima-doctor/.
• The runtime is the source of truth for diagnosis and repair progress.
• Never ask the user to paste curl | bash, iex (irm ...), base64 payloads, or other remote shell one-liners.
• If the runtime is not installed, direct the user to download it from the GitHub release: https://github.com/Approaching-AI/aima-openclaw/releases and extract into ~/.openclaw/tools/aima-doctor/. Verify the archive checksum: sha256:0f38959c6759e72a25bf089c9ab4aef1cf0022ba3aef1ea6686593896e973d1e.

Execution rules

1. Before running the helper, check that the runtime exists. On macOS/Linux check for ~/.openclaw/tools/aima-doctor/run.sh; on Windows check for %USERPROFILE%\.openclaw\tools\aima-doctor\run.ps1.
2. If the runtime is missing, install it automatically:
   • Download aima-doctor-runtime.zip from https://github.com/Approaching-AI/aima-openclaw/releases/download/v0.2/aima-doctor-runtime.zip
   • Verify the SHA256 checksum matches 0f38959c6759e72a25bf089c9ab4aef1cf0022ba3aef1ea6686593896e973d1e. If it does not match, stop and warn the user.
   • Create the target directory ~/.openclaw/tools/aima-doctor/ if it does not exist.
   • Extract the archive into that directory.
   • Confirm the runtime is now present before proceeding.
3. Prefer the packaged helper over ad-hoc shell commands.
4. Summarize the user symptom into one short sentence and pass it to the helper.
5. On macOS or Linux, run:

bash ~/.openclaw/tools/aima-doctor/run.sh --run --terminal --symptom "<symptom>"

6. On Windows, run:

& "$env:USERPROFILE\.openclaw\tools\aima-doctor\run.ps1" --run --terminal --symptom "<symptom>"

7. If the helper asks a question, relay that question to the user and rerun or continue with the answer as required.
8. If the helper reports success or failure, quote that result plainly. Do not invent a better outcome.
9. If the automatic download in step 2 fails, direct the user to the GitHub release at https://github.com/Approaching-AI/aima-openclaw/releases to download and extract manually.

Scope

• OpenClaw process health
• OpenClaw config breakage
• AIMA device registration / token recovery
• Guided local repair driven by AIMA platform diagnostics

Safety

• Treat helper output as operational evidence.
• Keep sensitive tokens and recovery codes out of normal chat unless the user explicitly provides them for repair.