SafeAI — Static AI Capability & Risk Analyzer

SafeAI is a static analysis tool that scans AI application source code for security risks, capability exposure, and governance gaps. It runs entirely offline, never executes agents or calls LLMs, and integrates into CI/CD pipelines.

🌐 safeai-analyzer.ikaruscareer.com — project landing page

Know Your Agent (KYA)

SafeAI now turns static scan results into a private, historical inventory of AI agents and their findings.

Why SafeAI?

Traditional application security tools (SAST, SCA, IaC scanning) are not designed for AI agent systems. AI applications introduce new risk surfaces:

• Prompt injection — untrusted input flows into model prompts
• Agent tool misuse — agents with filesystem, shell, or database access
• Capability sprawl — frameworks expose capabilities without visibility
• MCP exposure — Model Context Protocol endpoints and tools
• Governance gaps — missing authentication, permissions, audit trails

SafeAI fills this gap by analyzing frameworks, agents, tools, capabilities, and MCP integrations at rest—before deployment.

SafeAI analyzes AI applications without executing them, helping developers discover capabilities, identify potential risks, and improve governance early in the software lifecycle.

Designed to be lightweight, explainable, and community-driven, SafeAI aims to become an open foundation for AI capability and risk analysis.

SafeAI sits before runtime guardrails and red-teaming tools in the security lifecycle. It scans agent source code at commit time — detecting framework-specific capabilities, MCP misconfigurations, and prompt injection patterns — before you ever deploy an agent to staging. It does not replace runtime tools (Microsoft AGT), evaluation frameworks (LangSmith, DeepEval), or red-teaming scanners (Promptfoo, Garak). It complements them: find the risk in code first, then validate at runtime.

Key Features

How It Works

Source Code
    │
    ▼
File Collection — Python, YAML, JSON, .prompt, and .claude configs;
                 prunes VCS, caches, oversized files, and SafeAI's own artifacts
    │
    ▼
Framework Detection — 15 parsers (AST + config + regex), all run on all files;
                     import graph and dependency manifests
    │
    ▼
Static Analysis — semantic docs, component extraction, capability / prompt /
                 data-leakage / MCP / Claude Code analyzers
    │
    ▼
Capability Mapping — per-tool identity (agent, MCP server, skill, tool,
                    workflow node) + access modes (read < write < mutate < execute)
    │
    ▼
Risk Rules — rule engine with severity, confidence, provenance, stable fingerprints
    │
    ▼
Trust Score — deterministic 0–100 score across 7 weighted risk categories
    │
    ▼
KYA Pipeline — finding normalization, suppressions, baseline (new/regressed),
              policy-as-code, capability escalation diff
    │
    ▼
Registry & Reports — shared SQLite registry; terminal, JSON, SARIF 2.1.0, HTML,
                     canonical manifest, PR comment

Supported Frameworks

Framework Support Details

• LangGraph — detects StateGraph, add_edge, bind_tools, nodes, models
• CrewAI — detects Agent, Task, tools, models
• LangChain — detects AgentExecutor, Chain, Tool, PromptTemplate, models
• Semantic Kernel — detects Kernel.invoke, plugins, functions, skills, memory
• OpenAI Agents SDK — detects Agent, tools, handoffs, MCP references
• Microsoft Agent Framework — detects AgentClient, tools, workflows, Azure models
• Azure AI Foundry — detects YAML configurations with Azure resources
• Bedrock Agent — detects JSON configurations with Bedrock resources
• Claude Code — structural analysis of .claude/settings.json, permission grants, .mcp.json, slash commands, subagent definitions, and lifecycle hooks
• Google ADK — detects ADK agent, workflow, tool, and model patterns
• Mastra — detects Mastra agents, workflows, tools, and model references
• Haystack — detects Haystack pipelines, agents, tools, and retrievers
• LlamaIndex — detects agents, tools, indexes, and model references
• Dify — detects Dify workflow and agent configuration files
• n8n — detects n8n workflow exports, nodes, and connections

Maturity is on the scale defined in FRAMEWORK_SUPPORT.md: Partial = reliable detection and discovery with capability/risk analysis over common patterns; Experimental = detection and basic artifact discovery with limited framework-specific analysis. No framework is rated fully Supported yet — SafeAI is in early preview and deliberately does not overclaim coverage.

Supported Capabilities

SafeAI fingerprints capabilities at the framework object level and via fallback regex patterns. Each capability includes evidence, confidence score, resolved definition, and provenance.

Note: A capability is detected wherever the evidence lives — through a framework adapter, a direct pattern detector (for example Docker, Kubernetes, S3, Slack, Jira, browser automation, GCP), or MCP configuration analysis. Capabilities that only MCP configuration exposes today (e.g. email, human approval gates) are still flagged — the tool is reported with an unattributed identity rather than a guessed owner.

Know Your Agent (KYA) — Shared Registry

Every scan automatically builds a private "Know Your Agent" registry of scan-derived agent records — no server, no account, no network call, no source upload. Scans from every project accumulate in one shared SQLite database (SAFEAI_REGISTRY env var or ~/.safeai/registry.db), so safeai registry list shows the whole organization's agents from any folder.

safeai scan .                              # scan + accumulate into the shared registry
safeai scan . --manifest safeai-manifest.json   # also write the canonical KYA manifest
safeai scan . --html report.html                # interactive HTML report (risk gauge, escalations)
safeai registry list                       # agents/workflows from every scanned project
safeai registry list --format html > registry.html   # shareable HTML inventory
safeai registry show <agent-id>            # latest KYA record
safeai registry history <agent-id>         # all scans for an agent
safeai registry diff <agent-id> --from previous --to latest
safeai registry export --format json --output inventory.json
safeai registry export --format html --output inventory.html

What you get on the first run:

• A static scan ran successfully.
• The shared registry was initialized (SAFEAI_REGISTRY or ~/.safeai/registry.db).
• One or more KYA agent records were created with stable identities.
• Findings carry confidence, provenance, remediation, and stable fingerprints.
• No source code or secrets are uploaded or stored in output artifacts.

KYA records static evidence, not runtime truth. It answers "what does the source/configuration say this agent can do?" — never "what is this agent doing in production?" See REGISTRY.md, KYA_MANIFEST.md, and LIMITATIONS.md.

CI note: registry persistence is auto-disabled for bare CI jobs (the CI env var). Use --registry "$RUNNER_TEMP/registry.db", set SAFEAI_REGISTRY to a shared path, or use --no-registry for ephemeral scans.

Installation

Requirements

• Python 3.11, 3.12, or 3.13
• PyYAML (for YAML configuration parsing)

Install from source

git clone https://github.com/ikaruscareer/SafeAI.git
cd SafeAI
pip install -e .

Install development dependencies

pip install -e ".[dev]"

CLI Usage

python -m safeai scan <directory> [options]

python -m safeai registry <subcommand> [options]

Options

Exit Codes

Suppressed findings never trigger exit code 1. With --fail-on-new, only findings classified new or regressed against the baseline are gated.

Common 1.4 Workflows

# canonical manifest + baseline seed
python -m safeai scan . --manifest safeai-manifest.json

# CI/PR scan: fail only for new or regressed findings
python -m safeai scan . --baseline safeai-manifest.json --fail-on-new --fail-on high

# CI/PR scan: fail on capability escalations and render a PR comment
python -m safeai scan . --baseline safeai-manifest.json \
  --fail-on-escalation high --pr-comment comment.md

# inspect the shared KYA registry
python -m safeai registry list
python -m safeai registry show <agent-id>
python -m safeai registry history <agent-id>
python -m safeai registry diff <agent-id> --from previous --to latest
python -m safeai registry export --format json --output safeai-kya-inventory.json

Example Output

Terminal

SafeAI Scan Summary
Files: 12
Frameworks: langgraph, crewai
MCP assets: 2
Overall AI Risk Score: 73
critical: 1
high: 3
medium: 5
Findings:
[critical] app.py:10 - Untrusted input interpolated into prompt
[high] app.py:22 - Capability detected: shell_execution
[high] mcp.json:1 - MCP configuration does not define authentication

Example: LangGraph agent with MCP

{
  "Framework": "LangGraph",
  "Capabilities": ["Planner", "Memory", "Filesystem", "MCP"],
  "Risk Score": 73,
  "Findings": 9,
  "Critical": 1,
  "High": 3
}

CI/CD Integration

GitHub Actions

A workflow is included at .github/workflows/ci.yml. To use in your project:

jobs:
  safeai-scan:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: '3.12'
      - name: Install SafeAI
        run: |
          pip install -e .
      - name: Run scan
        run: |
          python -m safeai scan . \
            --sarif results.sarif \
            --html report.html \
            --manifest safeai-manifest.json \
            --no-registry
      - name: Upload SARIF
        uses: github/codeql-action/upload-sarif@v3
        with:
          sarif_file: results.sarif

GitLab CI

safeai-scan:
  image: python:3.12
  script:
    - pip install -e .
    - python -m safeai scan . --sarif results.sarif --html report.html --no-registry
  artifacts:
    paths:
      - results.sarif
      - report.html

Azure DevOps

- task: PythonScript@0
  inputs:
    scriptSource: 'inline'
    script: |
      import subprocess
      subprocess.run(["pip", "install", "-e", "."])
      subprocess.run(["python", "-m", "safeai", "scan", ".", "--sarif", "$(Build.ArtifactStagingDirectory)/results.sarif", "--no-registry"])

Capability escalation in CI

A capability escalation is a change between two scans where a tool gains more authority than it had before — a new shell capability, a filesystem access widening from read to write, a new MCP server, an approval gate being removed, and so on (see RULES_REFERENCE.md and KYA_MANIFEST.md for the full rule list). Reviewing these on every pull request is more targeted than reviewing every finding, because most findings on a mature codebase are pre-existing and already accepted; an escalation is new by definition.

--fail-on-escalation gates the scan on escalation severity, and --pr-comment writes a short Markdown summary you can post as a PR comment. SafeAI itself never posts anything anywhere and makes no network calls of any kind — generating the comment and publishing it are two separate steps, and the second one is entirely up to your CI workflow.

name: safeai-escalation-check
on:
  pull_request:

permissions:
  contents: read
  pull-requests: write

jobs:
  safeai-scan:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: '3.12'
      - name: Install SafeAI
        run: pip install -e .

      - name: Fetch baseline manifest from the base branch
        run: |
          git fetch origin "${{ github.event.pull_request.base.ref }}" --depth=1
          git show "origin/${{ github.event.pull_request.base.ref }}:safeai-manifest.json" \
            > safeai-manifest.json || echo '{}' > safeai-manifest.json

      - name: Run scan
        run: |
          safeai scan . \
            --baseline safeai-manifest.json \
            --pr-comment comment.md \
            --fail-on-escalation high

      - name: Post or update PR comment
        if: always()
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        run: |
          gh pr comment "${{ github.event.pull_request.number }}" \
            --edit-last --body-file comment.md \
            || gh pr comment "${{ github.event.pull_request.number }}" \
            --body-file comment.md

The gh pr comment --edit-last call updates SafeAI's own previous comment in place on repeat pushes, rather than adding a new one each time; it fails when there is no previous comment to edit (for example, on the first push), so the fallback plain gh pr comment handles that case. The --fail-on-escalation step runs before the comment step so the workflow's exit code still reflects the scan outcome; if: always() on the comment step makes sure the comment is posted even when the scan step fails the job.

SARIF Integration

SafeAI outputs SARIF 2.1.0 format, compatible with GitHub Advanced Security, Azure DevOps, and other SARIF-compliant tools.

Roadmap

See ROADMAP.md for the detailed roadmap.

• Completed in 1.3: KYA manifest, baseline/new-regressed gating, suppressions, policy-as-code, local SQLite registry, registry CLI.
• Completed in 1.4 (beta): tool-centric capability model (tool identity
  • access modes), 14 capability escalation rules, capability diff v2, deep Claude Code analysis, PR comment + CI context, assurance boundary, registry schema v2, shared org-wide registry default.
• Next focus: adapter depth improvements, governance signal detection, richer dataflow/context precision, and optional enterprise-scale workflows.

License

SafeAI is released under the Apache 2.0 License.