zed-agent-diagnostics
Skill319 starsChanged 11 days agoMITView on GitHub
When running as Zed's built-in coding agent in JETLS, prefer the faster built-in diagnostics tool for project instructions that require `jetls check` or `./scripts/selfcheck.sh`, with fallback to selfcheck when needed.
In this repository7 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- changelogSkill
- codex-review-loopSkill
- commitSkill
- run-testSkill
- write-testSkill
The file
We split this skill into the rules it gives an agent
9 rules. About 0 are common, 5 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 11
Use this skill when you are running as Zed's built-in coding agent in this repository and need to satisfy project instructions that require …
- Very specific: line 17
When project instructions say to run jetls check or ./scripts/selfcheck.sh after modifying code, normally satisfy that requirement by using …
- Very specific: line 22
Use Zed diagnostics for the initial check.
- Distinctive: line 27
After code edits, call the diagnostics tool for the project.
- Distinctive: line 28
If the changed files are known and targeted diagnostics are useful, also call the diagnostics tool for those files.
- Distinctive: line 30
Treat diagnostics in files you edited as issues to fix when they are likely caused by your changes.
- Distinctive: line 32
Do not remove or simplify meaningful code only to silence diagnostics.
- Distinctive: line 35
If diagnostics appear unrelated to your changes, report them as pre-existing or unrelated instead of fixing them opportunistically.
- Very specific: line 49
If diagnostics look stale, incomplete, or broken, do not spend time on server restart loops.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Zed agent diagnostics
Use this skill when you are running as Zed's built-in coding agent in this
repository and need to satisfy project instructions that require jetls check
or ./scripts/selfcheck.sh.
Validation rule
When project instructions say to run jetls check or ./scripts/selfcheck.sh
after modifying code, normally satisfy that requirement by using Zed's
diagnostics tool instead of running the terminal command. Prefer it when
available because it is usually faster than starting a separate selfcheck.
Use Zed diagnostics for the initial check. Run ./scripts/selfcheck.sh only
when diagnostics cannot provide enough confidence.
How to use diagnostics
- After code edits, call the diagnostics tool for the project.
- If the changed files are known and targeted diagnostics are useful, also call the diagnostics tool for those files.
- Treat diagnostics in files you edited as issues to fix when they are likely caused by your changes.
- Do not remove or simplify meaningful code only to silence diagnostics. If a natural fix is difficult to find, explain the situation and ask the user for help.
- If diagnostics appear unrelated to your changes, report them as pre-existing or unrelated instead of fixing them opportunistically.
When diagnostics may be unreliable
JETLS analyzes itself through a Revise-based server session. This is efficient, but changes to type definitions or global bindings may not be fully reflected in the running session. Zed diagnostics can therefore be stale, incomplete, or fail.
Be especially conservative when changes affect type definitions, macros,
include structure, diagnostics infrastructure, or code that diagnostics
itself depends on. If diagnostics code was edited, diagnostics may be broken
by the current changes.
If diagnostics look stale, incomplete, or broken, do not spend time on server
restart loops. Explain that Zed diagnostics do not provide enough confidence and
run ./scripts/selfcheck.sh to get a fresh check, unless the user has asked not
to run it.
Relationship to tests
This skill only changes how self diagnostics are checked. It does not replace relevant component-specific tests. When you modify testable code, still run the most specific appropriate tests as described by the project instructions.
Reporting
In the final response, say whether validation used Zed diagnostics, fell back
to ./scripts/selfcheck.sh, or used both. Summarize whether diagnostics were
clean or what issues remained.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.