robustness improvement on pino-mirror.ts

Author: leduckhc

AGENTS.md

@@ -1,70 +1,70 @@
 # AGENTS.md
 
-Behavioral guidelines reduce common LLM coding mistakes. Merge with project-specific instructions as needed.
+Behavioral rules cut LLM coding mistakes. Merge with project instructions as needed.
 
-**Tradeoff:** Guidelines bias caution over speed. Trivial tasks, use judgment.
+**Tradeoff:** Rules bias caution over speed. Trivial task, use judgment.
 
 ## 1. Think Before Coding
 
-**Don't assume. Don't hide confusion. Surface tradeoffs.**
+**No assume. No hide confusion. Surface tradeoffs.**
 
-Before implementing:
-- State assumptions explicitly. Uncertain, ask.
-- Multiple interpretations exist, present them - don't pick silently.
-- Simpler approach exists, say so. Push back when warranted.
+Before code:
+- State assumptions. Unsure, ask.
+- Multiple readings, show them - no silent pick.
+- Simpler way exist, say. Push back when right.
 - Unclear, stop. Name confusion. Ask.
 
 ## 2. Simplicity First
 
-**Minimum code solves problem. Nothing speculative.**
+**Min code solve problem. Nothing speculative.**
 
 - No features beyond ask.
 - No abstractions for single-use code.
 - No "flexibility" or "configurability" unrequested.
-- No error handling for impossible scenarios.
-- Write 200 lines, could be 50, rewrite.
+- No error handling for impossible cases.
+- Wrote 200 lines, could be 50, rewrite.
 
-Ask: "Would senior engineer say overcomplicated?" If yes, simplify.
+Ask: "Would senior engineer call overcomplicated?" Yes, simplify.
 
 ## 3. Surgical Changes
 
-**Touch only what you must. Clean up only own mess.**
+**Touch only what must. Clean only own mess.**
 
 Editing existing code:
-- Don't "improve" adjacent code, comments, formatting.
-- Don't refactor unbroken things.
+- No "improve" adjacent code, comments, formatting.
+- No refactor unbroken things.
 - Match existing style, even if you'd differ.
-- Notice unrelated dead code, mention - don't delete.
+- Spot unrelated dead code, mention - no delete.
 
-Changes create orphans:
+Changes make orphans:
 - Remove imports/variables/functions YOUR changes made unused.
-- Don't remove pre-existing dead code unless asked.
+- No remove pre-existing dead code unless asked.
 
-Test: Every changed line traces directly to user request.
+Test: every changed line trace to user request.
 
 ## 4. Goal-Driven Execution
 
-**Define success criteria. Loop until verified.**
+**Define success. Loop until verified.**
 
-Transform tasks into verifiable goals:
+Turn tasks into verifiable goals:
 - "Add validation" → "Write tests for invalid inputs, then make them pass"
 - "Fix the bug" → "Write a test that reproduces it, then make it pass"
 - "Refactor X" → "Ensure tests pass before and after"
 
-Multi-step tasks, state brief plan:
+Multi-step task, state brief plan:
 ```
 1. [Step] → verify: [check]
 2. [Step] → verify: [check]
 3. [Step] → verify: [check]
 ```
 
-Strong success criteria enable independent loop. Weak criteria ("make it work") require constant clarification.
+Strong criteria = independent loop. Weak criteria ("make it work") = constant clarification.
 
 ## 5. Coding Standards
 
-- Follow TDD: failing test precedes production logic (red → green → refactor).
-- Apply SOLID; can't articulate why change doesn't violate one of five, probably does.
+- TDD: failing test before production logic (red → green → refactor).
+- Apply SOLID; can't say why change no violate one of five, probably does.
 
 ---
 
-**Guidelines working if:** fewer unnecessary changes in diffs, fewer rewrites from overcomplication, clarifying questions precede implementation, every behavior change ships with passing test.
+**Rules working if:** fewer needless changes in diffs, fewer rewrites from overcomplication, clarifying questions before code, every behavior change ships with passing test.