Skip to content

check tool

Use check after the agent writes or changes Python, as a person runs xpd check. It lints files and folders and returns every issue with its rule, place, severity and fix. It changes no file, so a client can run it without asking.

check(paths=null, unsafe_fixes=false, options=null)
  • Changes no file
  • A second call with the same arguments changes nothing more
  • Stays inside the project

Tags: lint.

The server gives the agent this description of the tool.

Lint Python files for slow or wrong habits with pandas, NumPy, SciPy, scikit-learn, scikit-lego, MAPIE, Category Encoders, linearmodels, arch, pymoo, plotly, Apache Hamilton and Python itself. Changes no file.

Run it after you write or change Python. Fix every error. Fix a warning unless you can say why the code is right here, and don't silence rules. An issue whose fix is set can be fixed with the fix tool. A pandas/missing-<package> warning names a package the project lacks, so ask the user before you add the dependency. The rule tool explains any rule with examples.

Type: list of string or null. Optional, default null.

Files or folders inside the project root. Folders respect .gitignore. Default: the whole root.

Type: boolean. Optional, default false.

Count unsafe fixes, which may change behavior, in fixable too.

Type: object or null. Optional, default null.

Which rules run, and the project file that adds the project's own rules.

Field Type Required Default Description
options.rules list of string no Packages or public rule names.
options.severity "error", "warning", "info", "hint" or null no null Run only the rules of this severity and above.
options.config string or null no null A project file to read instead of the nearest sgconfig.yml, relative to the project root.

What xpd check found, after the fixes it wrote.

Field Type Required Default Description
fixed integer yes Number of fixes written to disk.
fixable integer yes Issues left that the fix tool fixes at the level asked for.
issues list of object yes
issues[].package string yes Module name of the rule's pack, such as pandas.
issues[].rule string yes Rule id, such as chained-assignment.
issues[].file string yes
issues[].line integer or null yes
issues[].column integer or null yes
issues[].severity "error", "warning", "info", "hint" yes
issues[].message string yes
issues[].fix "safe", "unsafe" or null yes How safe the rule's fix is, or null for no fix.

The call below lints one file for one rule, and counts the unsafe fix in fixable.

{"paths":["samples/bad.py"],"unsafe_fixes":true,"options":{"rules":["pandas/chained-assignment"]}}
{"fixed":0,"fixable":1,"issues":[{"package":"pandas","rule":"chained-assignment","file":"samples/bad.py","line":6,"column":5,"severity":"error","message":"Chained assignment writes to a copy under copy-on-write, so the frame keeps its old values. Assign to df.loc[i, \"a\"] instead.","fix":"unsafe"}]}

A path outside the project root fails before xpd runs.

../outside.py is outside the project root /tmp/project. Pass a path inside it.

A failure of the binary, such as a file it cannot read, comes back as a tool error with the command and its message. Issues are a result, not an error.