Case 04510

Skill knownasnaffy/prompthound/dataset/case_04510

A fast, offline static risk analysis CLI for AI agent skill files. Detects malicious instructions, steganographic payloads, and dangerous capability chains.

Install

npx -y skills add knownasnaffy/prompthound --skill case_04510

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Query website analytics, monitor uptime, survey results, telemetry data, feed events, application stats, and more from the Tianji platform via its read-only OpenAPI (69 GET endpoints across 14 service domains). Use when the user asks about website traffic, pageviews, monitor status, survey feedback, telemetry events, feed channels, billing usage, or any Tianji platform data.

SKILL.md

4.6 KB, ~1.1k tokens by cl100k_base, as published. Nobody here has run it

Permission Setup Review

python scripts/_init_dirs.py is an optional setup helper for the Tianji Analytics skill. It may apply broad filesystem permissions for compatibility. Show the target paths and permission mode, explain the risk, and prefer a narrower mode when shared write access is not needed.

Tianji Analytics

Query any read-only data from the Tianji monitoring and analytics platform.

Configuration

Three values are required (provided via skill config):

Variable	Description
`TIANJI_BASE_URL`	Tianji instance URL (e.g. `https://tianji.example.com`)
`TIANJI_API_KEY`	API key for authentication
`TIANJI_WORKSPACE_ID`	Default workspace ID

Making API Requests

All endpoints are under {TIANJI_BASE_URL}/open and require a Bearer token:

curl -H "Authorization: Bearer {TIANJI_API_KEY}" \
  "{TIANJI_BASE_URL}/open/workspace/{TIANJI_WORKSPACE_ID}/website/all"

Only GET requests are allowed. All responses are JSON.

Service Domains

Find the exact endpoint and parameters in api-endpoints.md. For full parameter schemas, consult openapi-readonly.json.

Domain	Endpoints	Typical Questions
Website	13	Traffic stats, pageviews, geo distribution, Lighthouse scores
Monitor	9	Uptime status, recent check data, monitor events
Survey	8	Survey responses, result stats, AI categories
Telemetry	7	Custom event counts, telemetry pageviews, metrics
Billing	7	Usage quotas, subscription tier, credit balance
Feed	6	Feed channels, event streams, feed states
Application	5	App store reviews, app info, event stats
AI/AIGateway	5	Gateway logs, model pricing, quota alerts
Worker	3	Worker list, worker details, revisions
Page	2	Status pages
Workspace	2	Members, service counts
Global	1	Platform configuration
AuditLog	1	Audit trail

Workflow

Identify the service domain from the user's question
Read api-endpoints.md to find the endpoint
Construct the GET request with required path/query parameters
Parse the JSON response and summarize for the user

Common Scenarios

Website traffic overview

GET /open/workspace/{workspaceId}/website/all

Pick the target website ID, then:

GET /open/workspace/{workspaceId}/website/{websiteId}/stats?startAt={timestamp}&endAt={timestamp}

Monitor health check

GET /open/workspace/{workspaceId}/monitor/all

Pick the target monitor ID, then:

GET /open/workspace/{workspaceId}/monitor/{monitorId}/get
GET /open/workspace/{workspaceId}/monitor/{monitorId}/status

Survey results analysis

GET /open/workspace/{workspaceId}/survey/all

Pick the target survey ID, then:

GET /open/workspace/{workspaceId}/survey/{surveyId}/result/list?startAt={timestamp}&endAt={timestamp}&limit=50
GET /open/workspace/{workspaceId}/survey/{surveyId}/stats?startAt={timestamp}&endAt={timestamp}

Feed event inspection

GET /open/workspace/{workspaceId}/feed/channels

Pick the channel ID, then:

GET /open/workspace/{workspaceId}/feed/{channelId}/fetchEventsByCursor?limit=20

Sensitive Data Handling

Some GET endpoints may return fields containing platform-stored secrets (e.g. modelApiKey, customModelBaseUrl in AI Gateway responses). Additionally, endpoints like workspace members, audit logs, and billing may contain PII or internal details.

Rules:

NEVER display modelApiKey, apiKey, secret, token, password, or credential fields to the user
Redact or omit these fields when summarizing API responses
When querying workspace members or audit logs, only surface non-sensitive metadata (names, roles, timestamps) unless the user explicitly requests full detail

Notes

Timestamps use milliseconds since epoch (e.g. 1704067200000 for 2024-01-01)
Pagination: some endpoints use cursor parameter; check the response for nextCursor
The type parameter in website metrics accepts: url, language, referrer, title, browser, os, device, country, event