AgentConfig Audit MCP and rules scanner
Developers configuring Model Context Protocol clients and AI coding assistants risk leaking sensitive credentials, granting unbounded filesystem access, or exposing workspace instruction…
After payment, Stripe emails your download link within a few minutes (valid 7 days). Pass buyers also get the private All-Access Vault on GitHub.
Digital goods, delivered instantly. If a tool doesn't work as described, we fix it or refund you — see Terms.
requirements.txt — list of dependencies
What's in the ZIP
- Python source, README and requirements
Why PyForge: Ready-made tools with full source code.
More details
Developers configuring Model Context Protocol clients and AI coding assistants risk leaking sensitive credentials, granting unbounded filesystem access, or exposing workspace instruction files to hidden Unicode and prompt injection attacks. Standard static analysis tools overlook these ecosystem-specific configuration pitfalls across local editor configurations and agent rule files. The scanner inspects standard MCP configuration files and workspace agent instruction rules for hardcoded API keys, private keys, and sensitive tokens. It identifies unpinned server executions using npx or uvx runners and warns against broad filesystem arguments targeting root or home directories. Additionally, it flags unauthenticated remote endpoints while detecting invisible Unicode characters, bidirectional text overrides, and prompt injection directive patterns. Example: Run python3 agentconfig_audit_mcp_and_rule.py --help to view all CLI options. Run python3 agentconfig_audit_mcp_and_rule.py to audit standard operating system paths. Run python3 agentconfig_audit_mcp_and_rule.py . to inspect the current workspace rules and local configs. This tool performs static auditing only and does not automatically rewrite files, sandbox running agent processes, or monitor live network traffic.
Quick Start
Runs on Python 3.9 or newer. Check requirements.txt in the ZIP and install what it lists with pip install -r requirements.txt.
1 — Install the dependencies
pip install -r requirements.txt2 — Run it
How agents & programs use it: Install its requirements in a virtual environment; scripts and AI agents can then run it as a subprocess.
Why developers buy from PyForge
- Save hours: ready-made, tested tools. Skip writing, debugging and maintaining it yourself.
- Ready in minutes: one file, drop it in your project and run it.
- Zero dependencies: no pip install chains, no supply-chain risk.
- Full source code: readable, auditable, yours to modify.
- Runs locally and offline: no account, no API key, no data leaves your machine.
- Pay once, own it: no subscription.
- Tested before release: every tool is checked before it goes on sale.
All-Access Pass
Get every PyForge tool with the All-Access Pass
33 tools today plus every future release, lifetime commercial license, one payment of $99.
This tool: $49 · All 33 tools: $99
See the All-Access Pass