linting-inline-skill
In-session linting delegate — discovers the language linter, runs it scoped to touched files only, applies gate semantics from verification-loop-skill, and fixes or reports findings inline. Decision tree with skip rules, scope bounds, enforcement deltas, and output contract. Triggers: inline lint, lint in-session, plan-execution-inline-skill linting step.
Install
npx skillscat add darellchua2/civiltekk-skills/linting-inline-skill Install via the SkillsCat registry.
Linting (inline)
You are executing the linting delegate's workflow in this session. The
subagent's isolation (fresh context, sandbox, tier model) is replaced by the
discipline below — advisory, but the gate's verdict depends on it.
Decision tree
- Skip check: no linter configured for the changed languages (no eslint/
ruff/checkstyle/dotnet-format configs, no manifest scripts) → report
"lint-inviable", stop. Never add a linter config unprompted. - Discover: linter + invocation from manifests/config per
language-linting-skill(Ruff for Python, ESLint for TS/JS, Checkstyle/
SpotBugs for Java, dotnet format for C#). Autoloaded by that skill where
available. - Scope (scoped-lint rule): lint ONLY the files in the current diff
(git diff --name-only --diff-filter=AM). Full-suite lint is the gate's
job, not yours. - Gate semantics per
verification-loop-skill: auto-fixable → fix and
re-run once; remaining findings → report by severity with file:line. - INCONCLUSIVE (linter crashes, config broken) → report it as such; never
declare green without a run.
Scope bounds
- Fixes touch only the linted files and only the flagged findings.
- No config changes (adding plugins, disabling rules) — that is a finding.
Enforcement deltas (vs linting-subagent)
| Subagent enforcement | Inline discipline (you) |
|---|---|
| Fresh context window | State the touched-file list before linting so scope drift is visible |
| Isolated fix sandbox | Fixes land in the working tree directly — list every file you changed in Output |
| Tier model | Same model as the caller — flag uncertainty instead of suppressing findings |
Output contract
Status: [success | partial | failed | inconclusive]
Output: linter(s) run + files linted + findings (fixed vs reported, file:line)
Summary: ≤3 sentences
Issues: unfixable findings, config problems, or "None"
Categories
Install
npx skillscat add darellchua2/civiltekk-skills/linting-inline-skill Install via the SkillsCat registry.