ai-nexusBLOCK
Structured AI-entity discovery, evidence, and identity layer for agents.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
Discover, rank, and compare AI models, datasets, papers, tools, and benchmarks. Updated daily, scored by the Free2AITools Nexus Index (FNI).
Website: free2aitools.com
What It Does
- Cross-source catalog — Models, datasets, papers, tools, and benchmarks from HuggingFace, GitHub, ArXiv, Ollama, Replicate, Civitai, and more
- FNI Ranking — 5-factor composite score (Semantic, Authority, Popularity, Recency, Quality) with full breakdown in every API response
- Hardware-aware filtering — filter models by VRAM, license, and task-category metadata (heuristic filters, not verified compatibility)
- Daily Updates — Automated pipeline refreshes all data daily
For Developers & AI Agents
Model Selection API
curl -s https://free2aitools.com/api/v1/select \
-H "Content-Type: application/json" \
-d '{"task":"text-generation","constraints":{"max_vram_gb":8}}'Returns FNI-ranked catalog entries with params, VRAM estimates, license, and a factual FNI factor summary (fni_summary) — not a fit verdict; the caller decides.
Model Comparison API
# , are id values taken from /api/v1/search results — substitute your own. curl "https://free2aitools.com/api/v1/compare?ids=,"
Side-by-side comparison with FNI factor decomposition. Obtain ids from the Search API first; see the Developer Docs for a runnable search-first flow.
Search API
curl "https://free2aitools.com/api/v1/search?q=code+generation&limit=5"
MCP Server
Add to Claude, Cursor, Windsurf, or any MCP-compatible client:
{
"mcpServers": {
"free2aitools": {
"url": "https://free2aitools.com/api/mcp"
}
}
}5 tools: free2aitools_search, free2aitools_rank, free2aitools_explain, free2aitools_select_model, free2aitools_compare.
Also listed on the [Smithery directory listing](https://s
fef76e956ee5OBSERVED · 2026-10-11Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add stream-aggregator-rust --env AES_CRYPTO_KEY=${AES_CRYPTO_KEY} --env ALGOLIA_ADMIN_KEY=${ALGOLIA_ADMIN_KEY} --env C4S2_AUTHORITY_CENSUS=${C4S2_AUTHORITY_CENSUS} --env GEMINI_API_KEY=${GEMINI_API_KEY} -- npx -y [email protected]{
"mcpServers": {
"stream-aggregator-rust": {
"command": "npx",
"args": [
"-y",
"[email protected]"
],
"env": {
"AES_CRYPTO_KEY": "${AES_CRYPTO_KEY}",
"ALGOLIA_ADMIN_KEY": "${ALGOLIA_ADMIN_KEY}",
"C4S2_AUTHORITY_CENSUS": "${C4S2_AUTHORITY_CENSUS}",
"GEMINI_API_KEY": "${GEMINI_API_KEY}"
}
}
}
}Exposed tools (11)
11 read · 0 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
AWQ | read | Activation-aware Weight Quantization is a technical... |
Loading... | read | Please wait while we fetch the latest model data. |
answer_directly | read | Answer the user directly from your own knowledge without calling any tool. |
free2aitools_compare | read | Compare two or more catalog entities across attributes, returning a per-attribute evidence-backed comparison. |
free2aitools_discover_models | read | Retrieve candidate AI models with FNI evidence for a task. Returns evidence |
free2aitools_explain | read | Explain an entity with its evidence chain (provenance and source_trail) by canonical id. |
free2aitools_rank | read | Return a ranked ordering of catalog entities for a task or category, with the ranking signal. |
free2aitools_search | read | Search the Free2AITools catalog of AI tools, models, datasets and papers by query; returns ranked entities with metadata. |
free2aitools_select_model | read | Recommend candidate models for stated requirements; returns candidates with supporting evidence. |
model_catalog_generic | read | Look up basic public model metadata (name, parameters, license, downloads) by keyword. |
web_search | read | General keyword web search; returns title + snippet + url for matching pages. |
Trust audit
BLOCKgrade F · trust 56/100 Do not install this without reading the findings. The audit found something that could harm you or your machine.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | FAIL |
| L1 | Static analysis of the code | WARN |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- declared (4 observation(s))
- Shell
- declared (1 observation(s))
- Dependencies
- not all pinned
- Secrets in source
- found
Findings (25)
robots.txt.js
check_compliance.cpython-314.pyc
entity-utils.js.tmp
url: `http://127.0.0.1:${addr.port}/`,const TOKEN = 'ZZTOPSECRETPAYLOAD9911';
const SECRET = 'TOP_SECRET_README_BODY_token_abcdef0123';
const ws = await exec({ tool: "web_search", arguments: { query: "license" } });const cat = await exec({ tool: "model_catalog_generic", arguments: { keyword: "llm" } });expect(() => new Function(body)).not.toThrow();
return new Function('return (' + fm.slice(open + 2, end + 1) + ');')();/// Replaces JS's sequential fs.readdir + crypto.createHash('md5') loop.return { size: o.body.length, etag: '"' + crypto.createHash('md5').update(o.body).digest('hex') + '-3"' };async headObject(key) { this.opLog.push(['headObject', key]); const o = this.objects.get(key); if (!o) throw this._notFound(); return { size: o.body.length, etag: '"' + crypto.createHash('md5').updateconst dataHash = crypto.createHash('md5').update(const combinedHash = crypto.createHash('md5').update(dataHash + logicHash).digest('hex');expect(raw).not.toMatch(/auth\.json|credentials\.json|id_rsa|\bsk-[a-z0-9]/i); // #23 no credential artifact
mkdir -p ../../output/images
find ../../data/images -name "*.jpg" | while read img; do
"../../../.github/workflows/sdk-npm-publish.yml",
rejectsSync(() => validateArchiveEntries([{ name: 'output/../../evil', type: 'file' }]), 'HANDOFF_UNSAFE_ARCHIVE');rejectsSync(() => validateArchiveEntries([{ name: 'output/link', type: 'symlink', linkTarget: '../../../etc' }]), 'HANDOFF_UNSAFE_ARCHIVE');return { classification: 'NONCRITICAL_NETWORK_WARNING', severity: 'WARNING', reason: `${sig.reason} ${statusOrFail}; CORS-blocked beacon preserved` };&& /telemetry signature|beacon/i.test(e.reason);
console.log('[SRS-2A] OBSERVATION: CF RUM beacon may be CORS-blocked in CI; a 429 burst may transiently rate-limit a same-origin critical asset. No product impact established.');/\/rum(\?|$|\/)/i, // Real-User-Monitoring beacon (cf-rum / analytics)
Gates applied: no_behavioural_pass.
fef76e956ee5full audit observations/trust-audit/mcp-server/mosesy5688-cell__ai-nexus.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-11 | fef76e956ee5 | BLOCK | F | 56 | first audit |
Questions
What is the ai-nexus MCP server?
Structured AI-entity discovery, evidence, and identity layer for agents.
What tools does ai-nexus expose?
11 in total: 11 read-only, 0 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is ai-nexus safe to connect to an agent?
No — not without reading the findings first. The audit graded it F (56/100) and found 1 critical or high issue in the source. Each one is listed on this page with the file and line it is on.
What credentials does ai-nexus need?
It reads AES_CRYPTO_KEY, ALGOLIA_ADMIN_KEY, C4S2_AUTHORITY_CENSUS, GEMINI_API_KEY, GITHUB_TOKEN, HF_TOKEN, HUGGINGFACE_TOKEN, IDENTITY_CLUSTER_MAX_PASSES, KAGGLE_KEY, R2_ACCESS_KEY, R2_ACCESS_KEY_ID and R2_SECRET_ACCESS_KEY from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does ai-nexus run?
It speaks streamable-http, so it runs as a service you connect to over the network. It is published on npm as stream-aggregator-rust at 1.0.0.
How current is this page?
The grade is for one exact copy of the source (fef76e956ee5), read on 2026-10-11. The repository is watched and re-audited when it changes.