🔎 Evidence browser

Browse the trust index

Search by skill, publisher, category, or trust summary — then use the runtime filters to find cards with live test evidence. The two main lanes are baseline safety checks first and deeper follow-on functionality checks after that.

⚙️ Filters · 2 active
✨ Quick picks
🏷 Categories · web-and-frontend-development

🧾 Evidence level: source-scanned means local source evidence; catalog-only means thinner metadata-first coverage.

🧪 Runtime status: cards can show only the baseline safety lane or the deeper follow-on functionality lane, depending on how far the skill got. Some cards now also surface how the skill behaved when clearly fake credentials were present.

📏 Depth cue: tells you whether the evidence stops at baseline checks, includes follow-on functionality checks, or includes richer fixture/example proof.

⏱ Freshness cue: tells you whether the latest runtime evidence is from the last 24 hours, the last 7 days, or is older and therefore less current.

🩺 Failure confidence: distinguishes a first seen failure from a repeated failure or a regression after an earlier pass, so not every red row means the same thing.

🧪 Fake-auth behavior: when available, this tells you whether a skill handled clearly fake credentials cleanly, needed real access to continue, or behaved badly around credential-like input.

Results

Showing 24 of 924 skills in the browsable catalog view · reviewed: no · category: web-and-frontend-development · sort: score
This snapshot is for the current page of results, not the whole filtered universe.
Browse hint: slices with zero failures plus some source-scanned or reviewed entries deserve more attention first; fresh runtime evidence helps too, because old clean receipts can still hide current drift.

mirage-proxy

chandika · vsource-scanned
53
overall

Install and configure mirage-proxy as a transparent PII/secrets filter for OpenClaw LLM API calls. Handles binary installation, provider config, auto-restart, and multi-model routing through the proxy.

Insufficient Evidenceconfidence: source evidencesource-scanned
+ 1 more
privileged capability
Take: Source-aware scan found higher-privilege capability areas (token, oauth), but that alone is not evidence of malicious behavior.
Decision cue: Decent evidence base — source-level signals are available, so inspect the receipts.

monarch-money

davideasaf · vsource-scanned
53
overall

TypeScript library and CLI for Monarch Money budget management. Search transactions by date/merchant/amount, update categories, list accounts and budgets, manage authentication. Use when user asks about Monarch Money transactions, wants to categorize spending, needs to find specific transactions, or wants to automate budget tasks.

High Riskconfidence: source evidencesource-scanned
+ 1 more
suspicious
Take: Potentially suspicious implementation signals detected: rm -rf, sudo , password.
Decision cue: Proceed carefully — suspicious signals matter more than capability surface alone.

multi-channel-engagement-agent

story91 · vsource-scanned
53
overall

Autonomous social media engagement across Twitter, Farcaster, and Moltbook. Fetches trending content, generates persona-driven contextual replies, and tracks state to prevent duplicates. Use when you want to engage with trending posts, reply to social media content, build audience through authentic engagement, or automate social presence across multiple platforms. Triggers on "engage on twitter", "farcaster engagement", "reply to trending", "social engagement bot", "multi-platform engagement", "autonomous social replies". Features include content filtering, mention tracking, webhook notifications, user blacklist/whitelist, analytics tracking, and quote tweet/recast support.

High Riskconfidence: source evidencesource-scanned
+ 1 more
suspicious
Take: Potentially suspicious implementation signals detected: eval(.
Decision cue: Proceed carefully — suspicious signals matter more than capability surface alone.

osint-investigator

cineglobe · vsource-scanned
53
overall

Deep OSINT (Open Source Intelligence) investigations. Use when the user wants to research, find, or investigate any person, place, organisation, username, domain, IP address, phone number, image, vehicle, or object using publicly available information. Triggers on phrases like "find information on", "investigate", "look up", "who is", "trace this", "dig into", "OSINT search", "background check", or any request to gather open-source intelligence about a target. Performs deep multi-source analysis across web search, social media, DNS/WHOIS, image search, maps, public records, and more — returning a structured intelligence report.

High Riskconfidence: source evidencesource-scanned
+ 1 more
suspicious
Take: Potentially suspicious implementation signals detected: password.
Decision cue: Proceed carefully — suspicious signals matter more than capability surface alone.

playwright-npx

mahone-bot · vsource-scanned
53
overall

Fast browser automation using Node.js scripts with Playwright (run via `node script.mjs`). Use for web scraping, screenshots, form automation, and any browser task requiring programmatic control. For simple page fetching without JavaScript execution, use web_fetch first. For interactive CLI browsing without writing code, use browser tool or playwright-cli. This skill is ideal when you need full control, custom logic, or reusable scripts.

High Riskconfidence: source evidencesource-scanned
+ 1 more
suspicious
Take: Potentially suspicious implementation signals detected: eval(, rm -rf, password.
Decision cue: Proceed carefully — suspicious signals matter more than capability surface alone.

query-dbpedia

kidehen · vsource-scanned
53
overall

Transform natural language questions into SPARQL queries for DBpedia and generate beautiful HTML results pages. Query the DBpedia knowledge graph using plain English prompts.

Insufficient Evidenceconfidence: source evidencesource-scanned
+ 1 more
privileged capability
Take: Source-aware scan found normal operational surface via environment, network, or shell-related references.
Decision cue: Decent evidence base — source-level signals are available, so inspect the receipts.

reveal-feedback

tolulopeayo · vsource-scanned
53
overall

Interact with Reveal feedback infrastructure to manage products, create review tasks, read AI-analyzed user feedback, get sentiment insights, view submissions, manage notifications, and register webhooks. Use when the user asks about product feedback, user reviews, testing tasks, sentiment analysis, top issues, review submissions, marketing videos, or anything related to their Reveal account.

Insufficient Evidenceconfidence: source evidencesource-scanned
+ 1 more
privileged capability
Take: Source-aware scan found normal operational surface via environment, network, or shell-related references.
Decision cue: Decent evidence base — source-level signals are available, so inspect the receipts.

reveal-reviewer

tolulopeayo · vsource-scanned
53
overall

Review products on Reveal as an AI agent reviewer. Browse available review tasks, navigate target websites using agent-browser, take screenshots, record observations, and submit structured feedback to earn rewards. Use when the user wants to review a product, test an app, submit feedback, check available review tasks, or earn rewards on Reveal. Requires agent-browser skill for website navigation.

Insufficient Evidenceconfidence: source evidencesource-scanned
+ 1 more
privileged capability
Take: Source-aware scan found normal operational surface via environment, network, or shell-related references.
Decision cue: Decent evidence base — source-level signals are available, so inspect the receipts.

rlm-controller

skywyze · vsource-scanned
53
overall

RLM-style long-context controller that treats inputs as external context, slices/peeks/searches, and spawns recursive subcalls with strict safety limits. Use for huge docs, dense logs, or repository-scale analysis.

High Riskconfidence: source evidencesource-scanned
+ 1 more
suspicious
Take: Potentially suspicious implementation signals detected: eval(, rm -rf, password.
Decision cue: Proceed carefully — suspicious signals matter more than capability surface alone.

rollhub-casino

rollhub-dev · vsource-scanned
53
overall

Play casino games (dice, coinflip, roulette) on Agent Casino with real cryptocurrency. Provably fair gambling API for AI agents. Use when the user wants to gamble, play dice, flip a coin, bet on roulette, or interact with Agent Casino at agent.rollhub.com. Supports 70+ crypto chains for deposits/withdrawals.

Insufficient Evidenceconfidence: source evidencesource-scanned
+ 1 more
privileged capability
Take: Source-aware scan found higher-privilege capability areas (wallet), but that alone is not evidence of malicious behavior.
Decision cue: Decent evidence base — source-level signals are available, so inspect the receipts.

s2g-workflow-engine

helmutsreinis · vsource-scanned
53
overall

Connect to S2G (s2g.run) visual workflow automation platform over WebSocket. Execute workflow nodes as tools — password generators, hash functions, date math, format converters, database queries, knowledge base, and any custom node. Use when asked to run S2G workflows, execute S2G nodes, connect to S2G, manage S2G workflows, or interact with the S2G platform API.

High Riskconfidence: source evidencesource-scanned
+ 1 more
suspicious
Take: Potentially suspicious implementation signals detected: password.
Decision cue: Proceed carefully — suspicious signals matter more than capability surface alone.

scamshield-verifier

marcodzano-lgtm · vsource-scanned
53
overall

The ultimate Web3 & OpenClaw security layer. Verifies if a repository, skill, or wallet address is malicious using the x402 API.

Insufficient Evidenceconfidence: source evidencesource-scanned
+ 1 more
privileged capability
Take: Source-aware scan found higher-privilege capability areas (wallet, private key), but that alone is not evidence of malicious behavior.
Decision cue: Decent evidence base — source-level signals are available, so inspect the receipts.

secucheck

jooneyp · vsource-scanned
53
overall

Comprehensive security audit for OpenClaw. Scans 7 domains (runtime, channels, agents, cron, skills, sessions, network), supports 3 expertise levels, context-aware analysis, and visual dashboard. Read-only with localized reports.

High Riskconfidence: source evidencesource-scanned
+ 1 more
suspicious
Take: Potentially suspicious implementation signals detected: rm -rf, sudo , password.
Decision cue: Proceed carefully — suspicious signals matter more than capability surface alone.

secure-shopper

moodykong · vsource-scanned
53
overall

Asynchronous shopping research + checkout using secure-autofill (1Password-backed browser filling) with results recorded to workspace artifacts.

Use Cautionconfidence: source evidencesource-scanned
+ 1 more
suspicious
Take: Potentially suspicious implementation signals detected: password.
Decision cue: Proceed carefully — suspicious signals matter more than capability surface alone.

skill-vettr

britrik · vsource-scanned
53
overall

Static analysis security scanner for third-party OpenClaw skills.

High Riskbaseline safety checks failed · 7/8confidence: source evidence
+ 2 more
source-scannedsuspicious
Runtime receipts + what failed2026-03-16 16:15 UTC
baseline-v3evidence depth: baseline checks onlytested recently: within 24 hoursfirst failed run seen for this lanefake-auth behavior: handled cleanlyexpectation failed, passed, handled fake credentials cleanlyoutput 452 Bartifacts 2worker oc-sandboxsource stage: fresh copysuite 2442 ms
🕵️ expected proof signal was missing
fake-auth behavior: handled cleanlyClearly fake credentials were exercised and handled normally.
RatioDaemon muttered: The runtime lane gave skill-vettr a chance to act normal. It declined and talked a big game, then missed its own proof signal.7/8 baseline-v3 checks passed before the stumble. The source-mount check is the part that made this interesting.
Observed: 11 /workspace/source-files.txt
Take: Potentially suspicious implementation signals detected: eval(, rm -rf, password.
Decision cue: Review first — baseline-v3 already found trouble.

smart-meme-generator

olisim02 · vsource-scanned
53
overall

AI-powered meme generator that creates perfect, context-aware memes from any topic or situation. Use when user wants to create memes, needs reaction images, wants to make jokes visual, or needs viral social media content. Automatically selects optimal meme templates and generates witty captions based on context.

Use Cautionconfidence: source evidencesource-scanned
+ 1 more
suspicious
Take: Potentially suspicious implementation signals detected: password.
Decision cue: Proceed carefully — suspicious signals matter more than capability surface alone.

tide-watch

chrisagiddings · vsource-scanned
53
overall

Proactive session capacity monitoring and management for OpenClaw. Prevents context window lockups by warning at configurable thresholds (75%, 85%, 90%, 95%), automatically backing up sessions before resets, and managing session resumption prompts. Use when working on long-running projects, managing multiple conversation channels (Discord, Telegram, webchat), or preventing lost work from full context windows. Includes CLI tools for capacity checks, cross-session dashboards, archive management, and session resumption. Supports any model or provider.

High Riskbaseline safety checks failed · 7/8confidence: source evidence
+ 2 more
source-scannedsuspicious
Runtime receipts + what failed2026-03-16 00:45 UTC
baseline-v3evidence depth: baseline checks onlytested recently: within 7 daysfirst failed run seen for this laneexpectation failed, passedoutput 591 Bartifacts 2worker oc-sandboxsource stage: fresh copysuite 2351 ms
🕵️ expected proof signal was missing
RatioDaemon muttered: tide-watch talked a big game, then missed its own proof signal, which is not ideal for a skill asking to be trusted.7/8 baseline-v3 checks passed before the stumble. The source-mount check is the part that made this interesting.
Observed: 12 /workspace/source-files.txt
Take: Potentially suspicious implementation signals detected: rm -rf, sudo , password.
Decision cue: Review first — baseline-v3 already found trouble.

transparency-log-auditor

andyxinweiminicloud · vsource-scanned
53
overall

Helps verify that skill signing events are recorded in an independently

Insufficient Evidenceconfidence: source evidencesource-scanned
+ 1 more
privileged capability
Take: Source-aware scan found normal operational surface via environment, network, or shell-related references.
Decision cue: Decent evidence base — source-level signals are available, so inspect the receipts.

unipile-linkedin

sudhanshu746 · vsource-scanned
53
overall

Interact with LinkedIn via Unipile API - send messages, view profiles, manage connections, create posts, react to content. Use when the user asks to message someone on LinkedIn, check LinkedIn messages, view LinkedIn profiles, send connection requests, create LinkedIn posts, or interact with LinkedIn content.

High Riskconfidence: source evidencesource-scanned
+ 1 more
suspicious
Take: Potentially suspicious implementation signals detected: password.
Decision cue: Proceed carefully — suspicious signals matter more than capability surface alone.

url-fetcher

johstracke · vsource-scanned
53
overall

Simple web content fetching without API keys or external dependencies. Uses Python stdlib (urllib) only. Features: fetch HTML/text from URLs, basic HTML to markdown conversion, path-validated file writes (blocks system paths), URL validation (blocks localhost/internal). Security: File writes use is_safe_path() to prevent malicious writes. Perfect for content aggregation, research collection, and web scraping without API costs or dependencies.

Insufficient Evidenceconfidence: source evidencesource-scanned
+ 1 more
privileged capability
Take: Source-aware scan found normal operational surface via environment, network, or shell-related references.
Decision cue: Decent evidence base — source-level signals are available, so inspect the receipts.

veille

romain-grosos · vsource-scanned
53
overall

RSS feed aggregator, deduplication engine, LLM scoring, and output dispatcher for OpenClaw agents. Use when: fetching recent articles from configured sources, filtering already-seen URLs, deduplicating by topic, scoring with LLM, dispatching digests to Telegram/email/Nextcloud/file. Enhanced by mail-client (email output) and nextcloud-files (cloud storage).

High Riskfollow-on functionality checks could not be fully tested · 0/1confidence: source evidence
+ 2 more
source-scannedsuspicious
Runtime receipts + what blocked setup2026-03-15 21:31 UTC
functionality-v2evidence depth: follow-on functionality checkstested recently: within 7 daysfailure repeated in more than one runregression after earlier passneeded an external service the test could not reachoutput 375 Bartifacts 1worker oc-sandboxsource stage: cache hitsuite 5621 msbaseline-v3 8/8
🌍 needed an outside service the test could not reach
RatioDaemon on this skillVeille is built for veille. Follow-on functionality checks currently show the test still cannot run cleanly after an earlier pass, the trust label is High Risk, and setup looks advanced.
Take: Potentially suspicious implementation signals detected: eval(, rm -rf, password.
Decision cue: Review first — functionality-v2 already found trouble.

vibe-ship

satoshi891102 · vsource-scanned
53
overall

Ship a complete web app from idea to public deployment in one session. Use when user says "build me an app", "ship this idea", "vibe code", "quick ship", "deploy this", or describes any consumer app, tool, or website they want built and deployed. Handles validation, tech stack selection, building, testing, and deployment to Vercel or GitHub Pages.

Insufficient Evidenceconfidence: source evidencesource-scanned
+ 1 more
privileged capability
Take: Source-aware scan found normal operational surface via environment, network, or shell-related references.
Decision cue: Decent evidence base — source-level signals are available, so inspect the receipts.

video-download

upupc · vsource-scanned
53
overall

Download videos from 1800+ websites and generate subtitles using Faster Whisper AI. Use when user wants to download videos from YouTube, Bilibili, Twitter, TikTok, Facebook, Vimeo, or any other supported video site, extract audio, or transcribe video content to text/subtitles.

Insufficient Evidenceconfidence: source evidencesource-scanned
+ 1 more
privileged capability
Take: Source-aware scan found higher-privilege capability areas (token, telegram), but that alone is not evidence of malicious behavior.
Decision cue: Decent evidence base — source-level signals are available, so inspect the receipts.

warren-deploy-mainnet

planetai87 · vsource-scanned
53
overall

Deploy websites and files permanently on MegaETH mainnet using SSTORE2. Agents use their own wallet and pay gas.

Use Cautionconfidence: source evidencesource-scanned
+ 1 more
privileged capability
Take: Source-aware scan found higher-privilege capability areas (wallet, private key, token), but that alone is not evidence of malicious behavior.
Decision cue: Decent evidence base — source-level signals are available, so inspect the receipts.