AI Sentinel·Part of Concordance Pro
Your teams are shipping faster with AI.
Is your governance keeping up?
Sentinel automatically detects AI-active repositories and scores them against the 10 SDLC standards that matter most when models are in the loop. No manual tagging. No surveys. Evidence from your own toolchain.
The problem
↑
PR volume goes up
AI tools lower the cost of writing code. More PRs get opened, reviewed, and merged — with less time per review.
↓
Review depth goes down
Higher volume means less attention per PR. Reviews become ceremonial. Prompt changes get no review at all.
⏸
Governance lags behind
CI pipelines, branch rules, ownership — all built for deterministic code. AI integration breaks the assumptions quietly.
Compliance platforms measure controls. Engineering metrics platforms measure delivery speed. Neither measures whether the underlying practices are keeping pace. That's the gap Sentinel was built to close.
How detection works
Sentinel scans four surfaces across your connected repositories. AI-accelerated teams are identified automatically — no configuration required.
📦SDK Dependencies
openai, anthropic, langchain, huggingface and 56+ others across npm, PyPI, Go, Maven, Cargo
⚙️AI Tooling Config
.cursor/, .coderabbit.yml, aider.conf, copilot-instructions.md
🔑CI API Keys
OPENAI_API_KEY, ANTHROPIC_API_KEY, and 13 other LLM provider env vars in workflow YAML
💬Prompt Artefacts
/prompts/, system_prompt.txt, *.prompt.md, AGENTS.md committed to the repo
What gets scored
10 standards from the Concordance Framework where AI raises the stakes. Sentinel scores all 10 automatically from your repository evidence.
2.2Architecture Decision RecordsDesignNo audit trail for AI adoption decisions
3.1Branch ProtectionDevelopmentAI-generated code bypassing human review
3.9Secrets ManagementDevelopmentLLM API keys exposed, larger blast radius
4.1CI PipelineTestingNo automated gate on prompt injection
5.8Feature FlaggingReleaseNo kill-switch without a full deployment
From scan to action in minutes
01
Detect
Concordance scans your connected repositories across four surfaces. AI-accelerated teams are flagged automatically — no manual tagging required.
02
Score
Each AI-accelerated team is scored across the 10 Velocity Governance standards using direct evidence from your repositories. No surveys, no self-assessment.
03
Surface
Gaps are ranked by consequence — Critical (score below 2.0) and High (below 2.5). Each finding includes specific remediation guidance, not generic advice.
04
Track
Re-scan next sprint. Governance improvement is measurable over time. Sentinel gives you a score you can move, not a checkbox you mark done.
Availability
Sentinel is part of Concordance Pro. The base assessment — GitHub or GitLab connected, scored across all 50 standards including the 10 Velocity Governance standards — is free for one team.
The full Sentinel portfolio view — AI detection across all teams, aggregate governance posture, trend tracking — is part of the Pro tier. See pricing.
See which of your teams are exposed.
Connect your SCM. Sentinel detects AI repos automatically. Free for one team.