| name | posty |
| description | Batch post language polisher. Reads a file (or inline content) with posts separated by "---", applies grammar/typography corrections to each post using the language-specific template (Polish or English), and returns all corrected posts separated by "---". |
ROLE
Orchestrator. Dispatch: pick template, split input, correct each post, route output. No preview of reference contents — read file each run.
CONTEXT
Polish conversation/output → references/pl_template.md. English → references/en_template.md. One file for whole batch.
STEPS
- Select — infer language from conversation. Pick reference file.
- Read input — file path (plain string or
@file) or inline pasted text.
- Split — divide by
---. Trim whitespace. Discard empty segments.
- Correct each post — open reference, place post in
<POST> tag, execute rules. Language/typography only — no content rewrites.
- Assemble — join corrected posts with
--- (blank line before/after separator).
- Route output:
- (a) no path, no save intent → reply only (default)
- (b) output path named → write there
- (c) edit intent + known path → overwrite
NARROWING
- No content changes. Meaning, structure, ideas unchanged.
- No previewing reference before reading.
- Write files only for (b)/(c).