joinly

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Security

Grade A, and why

joinly scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

.claude/settings.json · 29 lines

What it actually says

{
  "permissions": {
    "allow": [
      "Bash(uv run ruff *)",
      "Bash(uv run pyright *)",
      "Bash(git status*)",
      "Bash(git diff*)",
      "Bash(git log*)",
      "Bash(ls *)",
      "Bash(mkdir *)",
      "mcp__ide__getDiagnostics"
    ]
  },
  "hooks": {
    "PostToolUse": [
      {
        "matcher": "Edit|Write",
        "hooks": [
          {
            "type": "command",
            "command": "f=$(jq -r '.tool_input.file_path // empty'); [[ \"$f\" == *.py ]] && cd \"$CLAUDE_PROJECT_DIR\" && uv run ruff check --fix -q \"$f\" 2>/dev/null && uv run ruff format -q \"$f\" 2>/dev/null || true",
            "timeout": 10
          }
        ]
      }
    ]
  }
}

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

Related

Other settings, from other repositories