Writing index
Notes that return to a system.
Each essay starts in an existing README or accepted ADR and returns to the project that made the question concrete.
Topics
agent-systems 3 · evaluation 1 · testing 1 · llm 1 · fail-closed 2 · privacy-engineering 2 · pii-redaction 1 · mastra 1 · developer-tools 1 · retrieval 1 · sqlite 1 · fts5 1 · trust-boundaries 1 · local-first 1
Index
Every claim links to a public source.
Claims stay bounded to public project documentation and linked evidence.
An eval harness for a user-facing agent
An agent chatbot cannot be unit tested like a normal program. Career Copilot splits evaluation into a deterministic contract lane that blocks merges and a budgeted quality lane that never overrides it.
Drop-in PII redaction for Mastra agents
Agent loops touch sensitive text at every boundary — prompts, tool calls, traces, memory. The hard part is not detection; it is integration.
Separate the product surface from the benchmark harness
DocNeedle's README is persuasive because it treats normal CLI use and repo-specific evaluation as different things instead of collapsing them into one story.
Trusted ingress before more tools
Career Copilot shows why a single-owner agent benefits from tight trust boundaries before it grows new capabilities.