ttsc Blog
Engineering notes, release essays, and practical articles about TypeScript-Go, compiler plugins, typed execution, linting, and the ttsc toolchain.
Browse by tag: agents-md (2) · ai (4) · coding-agents (3) · eslint (1) · go (1) · llm (1) · mcp (1) · opensource (5) · spec-driven-development (2) · typescript (4)
Evidence Graph: 100% Coverage and Compliance in 20+ Languages
Modern AI writes well and follows badly. @wrtnlabs/evidence makes every rule, requirement, schema, and API an obligation the check enforces, in 20+ languages plus Markdown and...
Evidence Graph: Make Every SKILL Instruction 100% Enforced
Agents do not follow the rules you wrote in AGENTS.md. A checklist claim makes every function answer every rule, and a missing answer is a compile error.
I Made Evidence Graph: 100% Spec Coverage for Coding Agents
@ttsc/evidence turns missing specification obligations into compiler errors, replacing review-until-dry with a graph that reached 100% coverage on 4.8–13.3x fewer tokens.
I Made TS Compiler Graph MCP: 10x Fewer Tokens in Claude Code
Built on the TypeScript compiler's own APIs, it reads the graph the compiler already resolved instead of guessing at it, so it stays exact and cuts tokens about 10x even on...
800x Faster TypeScript Lint: I Moved Lint Into TypeScript-Go
@ttsc/lint moves lint into the TypeScript-Go-backed check path and ships a separate native format command: 800x faster lint and 70x faster format on the VS Code fixture.