# Filter Record

> Use when filtering records from Entrez/NCBI data streams as part of entrez-direct workflows.

- Skill: `vimalinx/filter-record` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add vimalinx/filter-record`
- Raw SKILL.md: https://api.skillmd.com/api/skills/vimalinx/filter-record/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: vimalinx (https://skillmd.com/u/vimalinx)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/vimalinx/filter-record

---


# filter-record

## Quick Start

- **Command:** `filter-record < records.txt > filtered.txt`
- **Local executable:** `/home/vimalinx/miniforge3/envs/bio/bin/filter-record`
- **Full reference:** See `references/help.md` for detailed usage

## When To Use This Tool

- Apply EDirect's built-in record-filtering step to a record-oriented text stream.
- Clean or narrow structured NCBI/Entrez records before a downstream transformation stage.
- Keep record filtering inside the `transmute` toolchain instead of building ad hoc shell parsing.

## Common Patterns

```bash
# 1) Filter a saved record stream
filter-record < records.txt > filtered.txt
```

```bash
# 2) Filter an upstream EDirect stream before the next pipeline stage
some-upstream-command | filter-record | some-downstream-command
```

## Recommended Workflow

1. Start from a representative upstream record stream.
2. Run `filter-record` in a pipe or with stdin redirection.
3. Inspect a small sample of the output before scaling up.
4. Continue into downstream EDirect steps once the retained record structure looks right.

## Guardrails

- This wrapper is just `transmute -txf`, so `transmute` must be on `PATH`.
- `--help` and `--version` are not implemented here; both are reported as unrecognized arguments.
- Prefer stdin or pipes over undocumented positional-file behavior.
- The real filtering semantics live inside `transmute`, so validate behavior on a representative record stream before batching.