ZacharyZhang-NY
e74dafcd36
Three independently ugly bugs that all fed the same log flood:
* Electron's `WebFrameMain.send` writes every "Render frame was disposed"
to stderr via `console.error`. When the worker fans a second
agent-event past a reloading renderer, we produced thousands of
identical lines and stalled the main loop on stderr I/O. Added a
`mainFrame.detached` guard so the send is short-circuited before
Electron's own catch block fires, and installed a targeted filter in
`suppress-warnings.ts` (now actually imported from the entry point)
so the specific `console.error("Error sending from webFrameMain:"…)`
noise never reaches the terminal. Any other payload still surfaces.
* `dispatchAgentMessage` accepted the human-reader sentinel as a wake
target when an agent replied to `__board__` — `wakeAgentIfPossible`
then called `db.getAgent("__board__")`, threw, and the dispatcher
kept repeating on every send. Strip `HUMAN_SENDER_ID` at
`resolveMessageWakeTargets`, and keep a defense-in-depth guard in the
heartbeat so a stray caller (future automation fan-out) cannot
reintroduce the crash.
* Pi's skills loader rejects any SKILL.md whose frontmatter `name`
diverges from its parent directory. The four bundled `my-writing-style`
copies declared `name: writing-style-skill`, so every install emitted
a warning and dropped the skill. They also lived as orphan gitlinks
(mode 160000 without .gitmodules), and the global `*.md` gitignore
hid every other SKILL.md from version control. Fixed the frontmatter
to match the dir, converted the dangling submodules to plain trees,
and allowlisted `builtin-skills/**/*.md` so a fresh clone carries the
catalogue instead of silently losing it.
1.9 KiB
1.9 KiB
title, impact, impactDescription, tags
| title | impact | impactDescription | tags |
|---|---|---|---|
| Write Learning Tests for Third-Party Code | MEDIUM | catches 90%+ of breaking changes in dependency upgrades before production | bound, boundary, learning-tests, third-party, testing |
Write Learning Tests for Third-Party Code
When integrating a third-party library, write small tests that verify your understanding of how the API works. These tests serve double duty: they help you learn the API, and they act as a safety net when upgrading the library.
Incorrect (learning by trial and error in production code):
// Experimenting directly in application code
public void importCustomerRecord(String json) {
// Does Gson handle null fields? Let's find out...
Gson gson = new Gson();
CustomerRecord record = gson.fromJson(json, CustomerRecord.class);
// Does this throw or return null for missing fields?
// Who knows — we'll find out in production
}
Correct (learning tests verify assumptions):
@Test
public void gsonShouldDeserializeNullFieldsAsNull() {
Gson gson = new Gson();
Data data = gson.fromJson("{}", Data.class);
assertNull(data.getName());
}
@Test
public void gsonShouldHandleMalformedJson() {
Gson gson = new Gson();
assertThrows(JsonSyntaxException.class,
() -> gson.fromJson("{invalid", Data.class));
}
@Test
public void gsonShouldIgnoreUnknownFields() {
Gson gson = new Gson();
Data data = gson.fromJson("{\"name\":\"test\",\"unknown\":true}", Data.class);
assertEquals("test", data.getName());
}
Benefits:
- Free education on the API's behavior
- When you upgrade the library, these tests break first — before your production code does
- Documents exactly which behaviors your code relies on
- Takes minutes to write but saves hours of debugging
Reference: Clean Code, Chapter 8: Boundaries