Seven structural checks on a public repository. Scanned 2026-11-02 23:18 · 1,837 files
55.2/ 100 · risk level High
The weakest of the three on process controls. No CODEOWNERS, no pre-commit hook, no CI gate semantics, no dependency lock, and test imports reach only 3 of the source packages. On a repository this size that combination means a change cannot be scoped before it lands.
Scorecard
#
Check
Score
Weight
Verdict
C1
Scope Containment
3/10
18%
needs fixing
C2
False-Green Detection
9/10
18%
pass
C3
Drift Containment
10/10
14%
pass
C4
Memory Integrity
4/10
10%
needs fixing
C5
Delivery Gate
6/10
16%
acceptable
C6
Complexity Control
2/10
12%
needs fixing
C7
Regression Surface
3/10
12%
needs fixing
Evidence
C1 Scope Containment3/10
CI config: .github/workflows
CODEOWNERS: none (no gate on critical paths)
Pre-commit hook: none
CI contains gate/guard/policy semantics: no
Risk: An agent can edit any file and push it with nothing in the way.
C2 False-Green Detection9/10
Source files 1177 / test files 303 (test ratio 25.7%)
Constraint strength: 22 hard prohibitions found (MUST / NEVER / forbidden)
C4 Memory Integrity4/10
Memory/state files: none found
Project shows agent/LLM characteristics: yes
Note: agent project without cross-session memory files. Within a single session this is not a conflict; it becomes a gap the moment state is expected to persist
Conflict arbitration: none (no arbiter when old and new context disagree)
Risk: No state anchor across sessions. The same question gets contradictory answers on different days.
Three largest files: python\packages\autogen-agentchat\tests\test_assistant_agent.py (3562 lines); python\packages\autogen-ext\tests\models\test_openai_model_client.py (3380 lines); python\packages\autogen-agentchat\tests\test_group_chat.py (1946 lines)
Risk: Oversized files. The agent's context truncates mid-edit and the regression lands unseen.
C7 Regression Surface3/10
Dependency lock: none (builds are not reproducible)
Source packages 12, reachable by test imports 0 (coverage 0%); not covered: agbench, autogen-agentchat, autogen-core, autogen-ext, autogen-studio, autogen-test-utils, component-schema-gen, docs
CI matrix across environments: yes
Risk: One change, no way to enumerate the blast radius. Regression becomes luck instead of a checklist.
Priority stack
#
Fix first
Score
Why
1
Complexity Control
2/10
Oversized files. The agent's context truncates mid-edit and the regression lands unseen.
2
Scope Containment
3/10
An agent can edit any file and push it with nothing in the way.
3
Regression Surface
3/10
One change, no way to enumerate the blast radius. Regression becomes luck instead of a checklist.
4
Memory Integrity
4/10
No state anchor across sessions. The same question gets contradictory answers on different days.
5
Delivery Gate
6/10
below threshold
Want this run on your repository?
A 7-day, human-verified audit: scored report, priority stack, and a recorded walkthrough of your own code. Findings are checked by hand and anything that does not hold up is deleted.