Security audit
Security checks for vulnerabilities and agentic risk
This is a Kimi API helper that clearly discloses sending prompts to Moonshot and only saves local workflow notes with user approval.
Install only if you intend to use Moonshot's Kimi API and are comfortable sending selected prompts there. Keep MOONSHOT_API_KEY in environment variables, do not paste keys into chats or logs, confirm before saving ~/kimi/ memory or logs, and redact sensitive customer, internal, or secret-bearing content before API calls.
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
integration: pending | done | declined
## Context
<!-- What the user is trying to achieve with Kimi -->
<!-- Example: Use Kimi as the cheap long-context pass before a stricter local validator -->
## Routes
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.
Before sending anything non-trivial to Kimi, verify:
1. Does the prompt contain secrets, access tokens, or internal hostnames?
2. Does it include customer identifiers or contract-sensitive text?
3. Can the task succeed with a redacted or metadata-only version?
4. Does the user want this exact content to leave the machine?
Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
- stores lightweight local notes only after user approval
This skill NEVER:
- invent live model availability without checking
- persist secrets in `~/kimi/`
- execute destructive downstream automation from unvalidated output
- treat cost-sensitive or sensitive-send boundaries as implicit
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
curl -s "${KIMI_BASE_URL}/chat/completions" \
-H "Authorization: Bearer $MOONSHOT_API_KEY" \
-H "Content-Type: application/json" \
-d '{
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
curl -s "${KIMI_BASE_URL}/chat/completions" \
-H "Authorization: Bearer $MOONSHOT_API_KEY" \
-H "Content-Type: application/json" \
-d '{
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
## Migration Order
1. Replace the base URL with `https://api.moonshot.ai/v1`.
2. Replace the auth env var lookup with `MOONSHOT_API_KEY`.
3. Refresh model IDs from `/models` instead of copying old names.
4. Re-run one minimal request before touching prompts or business logic.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
## Migration Order
1. Replace the base URL with `https://api.moonshot.ai/v1`.
2. Replace the auth env var lookup with `MOONSHOT_API_KEY`.
3. Refresh model IDs from `/models` instead of copying old names.
4. Re-run one minimal request before touching prompts or business logic.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
## Migration Order
1. Replace the base URL with `https://api.moonshot.ai/v1`.
2. Replace the auth env var lookup with `MOONSHOT_API_KEY`.
3. Refresh model IDs from `/models` instead of copying old names.
4. Re-run one minimal request before touching prompts or business logic.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
## Migration Order
1. Replace the base URL with `https://api.moonshot.ai/v1`.
2. Replace the auth env var lookup with `MOONSHOT_API_KEY`.
3. Refresh model IDs from `/models` instead of copying old names.
4. Re-run one minimal request before touching prompts or business logic.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
## Migration Order
1. Replace the base URL with `https://api.moonshot.ai/v1`.
2. Replace the auth env var lookup with `MOONSHOT_API_KEY`.
3. Refresh model IDs from `/models` instead of copying old names.
4. Re-run one minimal request before touching prompts or business logic.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
## Migration Order
1. Replace the base URL with `https://api.moonshot.ai/v1`.
2. Replace the auth env var lookup with `MOONSHOT_API_KEY`.
3. Refresh model IDs from `/models` instead of copying old names.
4. Re-run one minimal request before touching prompts or business logic.
The activation guidance is broad enough to trigger on generic mentions of 'Kimi', 'Moonshot', 'long-context analysis', or 'Kimi-based coding', which can cause the skill to engage outside the user's actual intent. In a routing and workflow skill, over-broad activation increases the chance of unsolicited persistence, unnecessary data handling prompts, or accidental steering of unrelated conversations into this skill's behavior.
The markdown instructs users to inspect and modify MOONSHOT_API_KEY, which is a sensitive credential. While the steps are operationally reasonable, the document does not warn users to avoid exposing, logging, or sharing the key while troubleshooting.
No suspicious patterns detected.