Comparison

coding-ethos overlaps with several policy, static-analysis, and agent-guidance tools, but it is not a drop-in replacement for any single one of them. Its role is to connect guidance and enforcement across human contributors, AI coding agents, Git hooks, CI, SARIF, MCP, and repo-local policy.

Short Version

Tool or pattern Primary strength How coding-ethos differs
pre-commit Runs configured hooks before commit Adds generated ETHOS docs, agent hooks, MCP, CEL policy, managed tool capture, SARIF, and runtime capability evidence.
CodeQL Deep semantic code analysis Focuses on repo workflow policy, agent behavior, generated configs, and custom guardrails around development actions.
Semgrep Pattern and semantic static analysis Uses existing linters plus CEL and Go evaluators to enforce repository workflow and agent-safety policies.
OPA/Rego General-purpose policy engine Keeps repo policy close to engineering principles, agent docs, hooks, MCP tools, and SARIF outputs.
Plain agent instructions Human-readable guidance for agents Compiles guidance into checked hook behavior, MCP responses, skills, axioms, and CI evidence.
GitHub branch protection Server-side merge gate Catches bad actions earlier in local hooks and agent tool-use hooks before the PR exists.

Compared With Pre-Commit

pre-commit is excellent at running hook commands. coding-ethos uses Git hooks too, but the core problem is broader: agents and humans need consistent guidance, policy decisions, lint remediation, and evidence across local tools, MCP, SARIF, and CI.

coding-ethos adds:

Compared With CodeQL

CodeQL is a deep static-analysis engine. coding-ethos is a repo-policy and agent-safety framework. They work well together: CodeQL can find semantic code risks, while coding-ethos can enforce development workflow constraints such as protected hook paths, agent memory writes, shell command safety, file growth, tool capability contracts, and generated config drift.

Compared With Semgrep

Semgrep is good for code-pattern rules. coding-ethos can consume findings from managed static-analysis tools, but it also evaluates proposed commands, file edits, Git actions, and tool runtime capabilities. CEL policies are intended for narrow repo-specific predicates where the input is already typed and policy-owned.

Compared With OPA

OPA is a general-purpose policy engine. coding-ethos uses CEL for custom policy because the immediate target is typed, local, fast hook enforcement with small repo-authored expressions. The larger design goal is not just policy evaluation; it is keeping policy, ETHOS principles, generated docs, remediation skills, MCP guidance, SARIF, and hooks aligned.

Compared With Agent Instructions Alone

Plain Markdown instructions are necessary but not sufficient. Agents can miss instructions, overfit to local context, or treat mismatched guidance and hooks as tool defects. coding-ethos still generates agent instructions, but it also enforces the same contract in tool-use hooks, Git hooks, CI, SARIF, MCP, and compiled policy bundles.

Where It Fits

Use coding-ethos when you want one source contract for: