GraqleBLOCK
Persistent organisational intelligence for AI agents. Turn codebases, documents, policies and decisions into a knowledge graph — 13 LLM backends, impact analysis, evidence + confidence. Works with Claude Code, Cursor, Copilot.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
Turn the context your organisation already has — codebases, documents, policies, decisions and workflows — into a persistent typed knowledge graph, so Claude Code, Cursor, Copilot and other agents reason over architecture, dependencies, prior lessons and evidence instead of re-reading disconnected files every session.
Models change. Tools change. Your architecture and institutional knowledge should not.
[](https://pypi.org/project/graqle/) [](https://python.org) []() []() []() [](#enterprise-trust) [](#patent--license)
pip install graqle
Website · Quickstart · How it works · Governed autonomy · Regulated deployments · Changelog · VS Code Extension
Why GraQle exists
Your organisation already knows the answer. It just can't hand it to an AI.
The refund threshold lives in a policy document. The reason you retain records for seven years is in an ADR nobody re-reads. The incident that
bfa211087e58OBSERVED · 2026-10-08Connect
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 graqle -- None graqle==0.85.0
Trust audit
BLOCKgrade F · trust 40/100 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 | PASS |
| L1 | Static analysis of the code | FAIL |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (6 observation(s))
- Network
- declared (7 observation(s))
- Shell
- declared (10 observation(s))
- Dependencies
- pinned
- Secrets in source
- found
Findings (25)
_p("pki", "rsa_private_key", r"-----BEGIN RSA PRIVATE KEY-----")_p("pki", "ec_private_key", r"-----BEGIN EC PRIVATE KEY-----")_p("pki", "pkcs8_private_key", r"-----BEGIN PRIVATE KEY-----")_p("pki", "openssh_private_key", r"-----BEGIN OPENSSH PRIVATE KEY-----")_p("ssh", "ssh_dsa_private", r"-----BEGIN DSA PRIVATE KEY-----")inst._model = pickle.loads(Path(path).read_bytes()) # noqa: S301
state: dict[str, Any] = pickle.load(fh) # noqa: S301
state: dict[str, Any] = pickle.load(fh) # noqa: S301
has_dangerous_target = any(k in code_text for k in ["sql", "cypher", "gremlin", "shell_exec", "os.system", "subprocess.run", "subprocess.call", "eval(", "exec("])results.append((PASS, "Key: AWS credentials", f"~/.aws/credentials: {masked}"))"169.254.169.254", # cloud instance metadata endpoint
"metadata.google.internal",
module = importlib.import_module(module_path)
mod = importlib.import_module(mod_name)
mod = importlib.import_module(mod_name)
importlib.import_module(pkg)
console.print(f" Key: {api_key[:8]}...{api_key[-4:]}" if len(api_key) > 12 else f" Key: {api_key}")console.print(f"[red]Cannot read token file {token}: {exc}[/red]")logger.error("SECURITY: rejected invalid stash token: %r", token[:50])logger.info("Rolled back via stash pop: %s", token)re.compile(r"\btheta_fold\b|\bθ_fold\b", re.IGNORECASE),
diff="DATABASE_URL = 'postgresql://user:secretpass@host:5432/db'",
text = "mongodb://admin:SuperSecret@host:27017/db"
text = "DATABASE_URL=mysql://root:password123@localhost:3306/mydb"
TOKEN = "TEST-MOCK-GATEWAY-TOKEN"
Gates applied: critical_finding, no_behavioural_pass.
bfa211087e58full audit observations/trust-audit/mcp-server/quantamixsol__graqle.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 |
|---|---|---|---|---|---|
| 2026-10-08 | bfa211087e58 | BLOCK | F | 40 | first audit |
Questions
What is the Graqle MCP server?
Persistent organisational intelligence for AI agents. Turn codebases, documents, policies and decisions into a knowledge graph — 13 LLM backends, impact analysis, evidence + confidence. Works with Claude Code, Cursor, Copilot.
Is Graqle safe to connect to an agent?
No — not without reading the findings first. The audit graded it F (40/100) and found 12 critical or high issues in the source. Each one is listed on this page with the file and line it is on.
What credentials does Graqle need?
It reads ANTHROPIC_API_KEY, API_KEY, AWS_ACCESS_KEY_ID, COGNIGRAPH_API_KEY, COGNIGRAPH_API_KEYS, COGNIGRAPH_LICENSE_KEY, DB_PASSWORD, GEMINI_API_KEY, GITHUB_TOKEN, GOOGLE_API_KEY, GOOGLE_APPLICATION_CREDENTIALS and GRAQLE_API_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 Graqle run?
It speaks stdio, so it runs as a local process your client starts. It is published on PyPI as graqle.
How current is this page?
The grade is for one exact copy of the source (bfa211087e58), read on 2026-10-08. The repository is watched and re-audited when it changes.