Automated summary

devpipe is a CLI that compresses text or code by dropping low-information tokens using LLM surprisal scoring, and generates technical specifications from a prompt through the Groq API.

Strengths

Local compression with a quantized Llama model via candle (Metal GPU on macOS), so text never leaves the machine
Flexible options: keep ratio, exact target token count, auto-detected optimal ratio, and compression stats
A generate-compress command that chains spec generation and compression in one call
Can write to a file or emit JSON, and can fall back to the Groq API for the compression step

Limitations

The first run downloads a roughly 700 MB GGUF model from HuggingFace
Spec generation requires a GROQ_API_KEY environment variable, hence network access and a Groq account
Installation assumes a specific local checkout (~/devpipe2) installed with cargo, and macOS 26+ may need an extra CXXFLAGS export

When to use it

Reach for it when you need to shrink a long document or source file below a token budget, or want a quick compressed technical spec drafted from a one-line description.

When not to use it

Avoid it when you cannot download a 700 MB model, have no network path to Groq for spec generation, or need a deterministic compressor with no language model involved.

Security analysis

Caution

Quality score85/100

The skill is a legitimate CLI wrapper, but it executes shell commands, downloads a large model from the network, and uses an API key for cloud inference. There is no destructive, obfuscated, or exfiltration behavior, so it warrants caution rather than a risky rating.

Findings

•Runs shell commands, including `cargo install --path .` from a hardcoded local directory `~/devpipe2`
•Downloads a ~700 MB GGUF model from HuggingFace on first run
•Requires `GROQ_API_KEY` and sends prompts to the Groq API, though it does not instruct exporting or transmitting secrets itself

How this analysis is made

Examples

Compress a file with stats

devpipe compress docs/architecture.md --keep-ratio 0.7 --stats

Auto-tuned compression to a token budget

Compress this file to at most 200 tokens and show me the stats: devpipe compress input.txt --target-tokens 200 --stats

Generate and compress a spec

devpipe generate-compress "Payment processing microservice" --auto --stats -o spec.md

devpipe

Name: Devpipe: text compression and spec generation
Author: Skills Guide Bot

Unified developer productivity CLI that compresses text by removing low-information tokens (LLM surprisal scoring) and generates technical specifications via Groq API.

Trigger

Use this skill when the user asks to:

Compress text or code to reduce token count
Remove low-information tokens from text
Generate a technical specification from a prompt
Generate a spec then compress the output

Instructions

Compress text

# From file
devpipe compress path/to/file.txt --keep-ratio 0.7 --stats

# From stdin
echo "long text" | devpipe compress --keep-ratio 0.7

# Auto-detect optimal keep ratio
devpipe compress input.txt --auto --stats

# Keep exact number of tokens
devpipe compress input.txt --target-tokens 200

# Use Groq API for compression instead of local model
devpipe compress input.txt --groq --stats

Generate a technical spec

Requires GROQ_API_KEY environment variable.

devpipe generate "A real-time collaborative editor" -o spec.md
devpipe generate "Auth system with OAuth2" --json --stats

Generate then compress

# Generate a spec and compress the output
devpipe generate-compress "Payment processing microservice" --stats

# With custom keep ratio
devpipe generate-compress "Auth system" --keep-ratio 0.5 --stats

# Auto-detect optimal compression ratio
devpipe generate-compress "CLI tool" --auto

# Use Groq API for compression step
devpipe generate-compress "API gateway" --groq

# Output to file as JSON
devpipe generate-compress "Auth system" -o spec.md --json

Installation

If devpipe is not installed:

cd ~/devpipe2 && cargo install --path .

Notes

Compression uses a local quantized Llama model via candle (Metal GPU on macOS).
First run downloads a ~700 MB GGUF model from HuggingFace (cached locally).
Spec generation uses Groq API (free tier, no credit card needed).
On macOS 26+, you may need: export CXXFLAGS="-isystem $(xcrun --show-sdk-path)/usr/include/c++/v1"

From the file to a first result

The download contains exactly the content displayed on this page. This copy may differ from the latest version in the source repository.

View the source and supporting files

This download contains SKILL.md only. Any scripts, references, templates and dependencies are not included. Get the required files from the source, preserving their directory structure, before continuing.

1. Download the file

Download SKILL.md

From the root of a test project, save the file as .skill-downloads/devpipe-text-compression-and-spec-generation-ms1aw4ro/SKILL.md, or use the command below. This folder is for inspecting the skill before installation.

Command for macOS/Linux or WSL: requires a POSIX shell and curl. On Windows without WSL, use the download button and create the folder manually.

mkdir -p '.skill-downloads/devpipe-text-compression-and-spec-generation-ms1aw4ro' && curl --fail --show-error --location 'https://skills-guides.com/api/skills/devpipe-text-compression-and-spec-generation-ms1aw4ro/download' --output '.skill-downloads/devpipe-text-compression-and-spec-generation-ms1aw4ro/SKILL.md.part' && mv '.skill-downloads/devpipe-text-compression-and-spec-generation-ms1aw4ro/SKILL.md.part' '.skill-downloads/devpipe-text-compression-and-spec-generation-ms1aw4ro/SKILL.md'

2. Check prerequisites

Read SKILL.md before using it. If it references scripts or external services, inspect them and prepare the required access in your test environment.

Original excerpt: Installation

If devpipe is not installed:

cd ~/devpipe2 && cargo install --path .

The platforms on this page are catalog indications. The paths below follow tool documentation; they do not prove this skill has been run in those tools. No minimum tool version is confirmed for this listing.

3. Load the skill

This file does not provide a usable name/description header for this native procedure. It remains available to download and inspect; use the manual method below.

Manual method or another tool: attach SKILL.md to a conversation, or open it in an agent that can read files. Include any required supporting resources. No universal folder is recognized by every tool.

4. Try a first task

Use a short, synthetic input: an API endpoint, a small document or an interface brief. In a new conversation, attach the file or give the agent its path, then adapt this request:

Read .skill-downloads/devpipe-text-compression-and-spec-generation-ms1aw4ro/SKILL.md and apply its instructions to this example: [describe a test input]. Identify any missing files, dependencies or access. Produce the result requested by the skill without inventing data or running unreviewed scripts.

Expected result: output matching the skill instructions, or a precise list of missing prerequisites. A read confirmation alone does not validate a first use. If an external resource is unavailable, that path remains untested.

Related skills

PDF Reader

Productivity

Read and extract text from PDF files, including documents, reports, contracts, and spreadsheets. Works with local files, URLs, and WhatsApp attachments. Use whenever PDF content is needed, even if not explicitly requested.

Claude CodeCursorCodexbeginnerSafe 20,122

Skills Guide Bot