Agent skill

Using Devql

by bitloops in bitloops/bitloops

A skill your agent uses when answering repo-understanding questions in a Bitloops repo with DevQL guidance enabled, especially when you need to locate code by path, line range, exact symbol, or…

Apache-2.0Auto-check passed

Install Using Devql

skills CLI
$ npx skills add bitloops/bitloops --skill using-devql -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install bitloops/bitloops using-devql --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/bitloops/bitloops.git skills-src && mkdir -p .claude/skills && cp -r skills-src/bitloops/src/host/hooks/augmentation/skills/using-devql .claude/skills/using-devql && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
using-devql
GitHub stars
243
Token cost
~1.6k tokens
SKILL.md length
719 words
Files
1
Skills in repo
2
Repo updated
First seen
Licence
Apache-2.0

At a glance

A skill your agent uses when answering repo-understanding questions in a Bitloops repo with DevQL guidance enabled, especially when you need to locate code by path, line range, exact symbol, or…

  • Works in 9 steps: Choose the most specific selector… → If the request is an identifier,… → If the request is approximate or… → …
  • Answering repo-understanding questions in a Bitloops repo with DevQL guidance enabled
  • SKILL.md covers Overview, When to Use, Choosing The by Selector and Process, plus 5 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Using Devql is an agent skill from bitloops/bitloops. Use when answering repo-understanding questions in a Bitloops repo with DevQL guidance enabled, especially when you need to locate code by path, line range, exact symbol, or approximate/conceptual search before reading files.

Its SKILL.md is about 1.6k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It works with Google Gemini and Rust. The repository describes itself as: Give AI coding agents the context they need to ship production-quality software. The licence is Apache-2.0.

When your agent uses it

  • Answering repo-understanding questions in a Bitloops repo with DevQL guidance enabled
  • Especially when you need to locate code by path
  • Approximate/conceptual search before reading files

Example prompts

  • “/using-devql”

Workflow steps

9 steps, taken from the first numbered list in SKILL.md.

  1. Choose the most specific selector available: path, path + lines, or
  2. If the request is an identifier, literal, path-like string, or snippet,
  3. If the request is approximate or conceptual, distill it to a short phrase
  4. Reserve searchMode: IDENTITY, ARCHITECTURE, CODE, or SUMMARY for
  5. Once you have a concrete file or artefact, run overview.
  6. If the response includes expandHint, schema, or another typed follow-up
  7. Expand only one relevant stage or artefacts(first: ...) if the overview
  8. When DevQL returns path, startLine, and endLine, use a bounded read
  9. If DevQL returns nothing useful, fall back to targeted repo search or file

What it can do on your machine

Read from SKILL.md and the folder at commit dc8980c. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md (its code samples are bash).

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Using Devql loads about 1.6k tokens when it runs. Until then it costs about 59 tokens; SKILL.md has 719 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~59
When it runs · the whole SKILL.md, loaded when a task matches
~1.6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from bitloops/bitloops at commit dc8980c, republished under its Apache-2.0 licence (© bitloops). 719 words, ~1,564 tokens.

Download SKILL.mdSave it as .claude/skills/using-devql/SKILL.md (or your agent's skills folder).
name
using-devql
description
Use when answering repo-understanding questions in a Bitloops repo with DevQL guidance enabled, especially when you need to locate code by path, line range, exact symbol, or approximate/conceptual search before reading files.

Using DevQL

Overview

DevQL is Bitloops's typed repo-intelligence surface. In repos where this guidance is enabled, use DevQL first for repo-understanding questions when it is available in the current session. If DevQL returns nothing useful or is not available, fall back to targeted repo search or file reads.

When to Use

  • understanding what a file, function, module, class, or symbol does
  • resolving the concrete artefacts matched by a path or line range
  • looking up an exact symbol with symbolFqn
  • looking up identifiers, literals, paths, or snippets with search plus searchMode: LEXICAL
  • looking up an approximate name or conceptual behavior with search in the default AUTO mode
  • answering architecture questions after selecting a concrete area

Choosing The by Selector

  • path: use this when the starting point is one file; Bitloops selects artefacts from that file.
  • path + lines: use this when the starting point is a specific region inside a file; Bitloops limits the seed artefacts to that line range.
  • symbolFqn: use this when the starting point is one exact artefact or symbol.
  • search: use this when you do not yet have an exact seed. Use searchMode: LEXICAL for identifiers, literals, file-ish strings, and snippets. Keep the default AUTO mode for approximate or conceptual search. Use IDENTITY, ARCHITECTURE, CODE, or SUMMARY only when you need advanced narrowing.

Process

  1. Choose the most specific selector available: path, path + lines, or symbolFqn when the request is exact; search when you need lookup rather than direct addressing.
  2. If the request is an identifier, literal, path-like string, or snippet, distill it and run search with searchMode: LEXICAL.
  3. If the request is approximate or conceptual, distill it to a short phrase and run search in the default AUTO mode first. Inspect artefacts(first: 10) and, when useful, searchBreakdown(first: 3) to see which retrieval mode is carrying the result.
  4. Reserve searchMode: IDENTITY, ARCHITECTURE, CODE, or SUMMARY for advanced narrowing when AUTO is broad but you already know which representation is likely to matter.
  5. Once you have a concrete file or artefact, run overview.
  6. If the response includes expandHint, schema, or another typed follow-up hint, read that hint before composing the next query.
  7. Expand only one relevant stage or artefacts(first: ...) if the overview shows that more detail is needed.
  8. When DevQL returns path, startLine, and endLine, use a bounded read around that range first: start about 20 lines before startLine and stop about 40 lines after endLine. Escalate to a full-file read only when the bounded context is insufficient.
  9. If DevQL returns nothing useful, fall back to targeted repo search or file reads.
Show full SKILL.md (288 more words)Show less

Query Templates

bash
# Use AUTO search first when the request is approximate or conceptual
bitloops devql query '{ selectArtefacts(by: { search: "<distilled conceptual phrase>" }) { count artefacts(first: 10) { path symbolFqn canonicalKind startLine endLine score summary } searchBreakdown(first: 3) { lexical { path symbolFqn score summary } identity { path symbolFqn score summary } code { path symbolFqn score summary } summary { path symbolFqn score summary } } } }'

# Use LEXICAL search for identifiers, literals, paths, or snippets
bitloops devql query '{ selectArtefacts(by: { search: "<identifier or snippet>", searchMode: LEXICAL }) { count artefacts(first: 10) { path symbolFqn canonicalKind startLine endLine score summary } } }'

# Ask for overview once the selection is concrete
bitloops devql query '{ selectArtefacts(by: { symbolFqn: "<symbol-fqn>" }) { overview } }'

# Same overview shape for file or file+line selectors
bitloops devql query '{ selectArtefacts(by: { path: "<repo-relative-path>", lines: { start: <start>, end: <end> } }) { overview } }'

Reading Response Hints

  • Treat overview as the compact first pass.
  • Treat searchBreakdown as an AUTO-only widening tool when the unified top hits are relevant but you want to see which lexical or semantic mode is also surfacing candidates.
  • When DevQL returns expandHint, schema, or another typed follow-up hint, read it before composing the next query.
  • Prefer the hint's suggested parameters and drill-down shape over guessing the next stage call.

Sandbox Execution

  • In sandboxed agent environments, run bitloops devql ... outside the sandbox by default.
  • These commands rely on Bitloops-managed daemon and runtime state under platform app directories, so they can fail inside a workspace-only sandbox even when DevQL is healthy.
  • If your platform requires approval or escalation for out-of-sandbox commands, request it immediately before running bitloops devql ....

Common Mistakes

  • Do not reintroduce obsolete selector names from older DevQL guidance.
  • Do not dump the full conversational prompt into search; distill it.
  • Do not use conceptual AUTO search for obvious identifiers or snippets when searchMode: LEXICAL is the right tool.
  • Do not force IDENTITY, CODE, or SUMMARY unless you are intentionally narrowing an already-understood search problem.
  • Do not ignore expandHint or other response hints when DevQL already tells you how to drill down.
  • Do not read an entire file first when DevQL already returned path, startLine, and endLine; start with a bounded read around the range, roughly 20 lines before startLine through 40 lines after endLine.
  • Do not start with stage-specific detail queries before selecting a concrete artefact.
  • Do not use DevQL for edits, builds, tests, formatting, or git operations.

Integration

  • If overview shows something relevant, expand only the one stage needed for the current question.
  • If DevQL is unavailable or empty, switch to targeted rg or file reads.

© bitloops, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in bitloops/src/host/hooks/augmentation/skills/using-devql of bitloops/bitloops.

Open the folder on GitHubat commit dc8980c

Compare with similar skills

Using Devql next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Using Devql compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Using Devql this skillbitloops/bitloops243—~1.6kAutomated safety check: PassApache-2.0
Ask Sonnerleepokai/Codync1774 repos~1.8kAutomated safety check: WarnMIT
Mobile Nativeleepokai/Codync1771 repos~4.2kAutomated safety check: PassMIT
Auto Improvecrimeacs/auto-improve135—~651Automated safety check: PassMIT
Run Shuntpleaseai/shunt264—~2.6kAutomated safety check: PassApache-2.0
Gridbash Panesjasonsuhari/gridbash134—~1.5kAutomated safety check: PassMIT

Similar skills

  • Ask Sonner

    leepokai/Codync

    Guide to Sonner, the React toast library — install and wire up the Toaster, pick the right toast() call, promise and loading toasts, updating, dismissing and persisting toasts, styling, theming and…

    177 GitHub starsUsed in 4 repos~1.8k tokens
    Frontend & DesignAuto-check: warnings
  • Mobile Native

    leepokai/Codync

    Make a web app feel native on a phone — the small CSS and meta-tag fixes that separate "a website in a browser" from something that feels installed.

    177 GitHub starsUsed in 1 repo~4.2k tokens
    MobileAuto-check passed
  • Auto Improve

    crimeacs/auto-improve

    GAN-style iterative improvement loop for any text artifact. An agent skill from crimeacs/auto-improve.

    135 GitHub stars~651 tokensUpdated 2 mo ago
    EducationAuto-check passed
  • Run Shunt

    pleaseai/shunt

    Build, launch, and drive shunt — the Claude Code LLM gateway (a Rust/axum Anthropic-Messages proxy).

    264 GitHub stars~2.6k tokensUpdated yesterday
    Testing & QAAuto-check passed
  • Gridbash Panes

    jasonsuhari/gridbash

    Coordinate with sibling agent panes from inside a GridBash grid — list panes and their roles, read what a pane is currently doing, prompt one pane or every other pane, and rename pane titles so the…

    134 GitHub stars~1.5k tokensUpdated yesterday
    DevelopmentAuto-check passed
  • Recall

    samzong/Recall

    Use Recall to search, inspect, continue, export, resume, or share indexed AI coding sessions.

    105 GitHub stars~3.7k tokensUpdated yesterday
    Auto-check passed

More from bitloops/bitloops

  • Devql Explore First

    bitloops/bitloops

    Use always before codebase exploration, bug fixing, SWE-bench tasks, locating symbols, files, or tests, or before using rg, grep, sed, find, or source-file reads in a repository.

    243 GitHub stars~897 tokensUpdated 4 mo ago
    Auto-check passed

Questions about Using Devql

What does Using Devql do?

A skill your agent uses when answering repo-understanding questions in a Bitloops repo with DevQL guidance enabled, especially when you need to locate code by path, line range, exact symbol, or…. Using Devql is an agent skill from bitloops/bitloops. Use when answering repo-understanding questions in a Bitloops repo with DevQL guidance enabled, especially when you need to locate code by path, line range, exact symbol, or approximate/conceptual search before reading files.

When should I use Using Devql?

Using Devql fits situations like: answering repo-understanding questions in a Bitloops repo with DevQL guidance enabled; especially when you need to locate code by path; approximate/conceptual search before reading files.

How do I install Using Devql in Claude Code?

Run `npx skills add bitloops/bitloops --skill using-devql -a claude-code`. Or copy the skill folder (bitloops/src/host/hooks/augmentation/skills/using-devql in bitloops/bitloops) into .claude/skills/using-devql in your project. Claude Code loads it when a task matches its description.

How do I install Using Devql in Codex?

Run `npx skills add bitloops/bitloops --skill using-devql -a codex`. Or copy the skill folder (bitloops/src/host/hooks/augmentation/skills/using-devql in bitloops/bitloops) into .agents/skills/using-devql in your project. Codex loads it when a task matches its description.

Can I use Using Devql in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add bitloops/bitloops --skill using-devql -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/using-devql, .gemini/skills/using-devql, .github/skills/using-devql and .opencode/skills/using-devql in your project.

What does Using Devql need to run?

SKILL.md names no scripts, command-line tools or credentials: Using Devql is instructions for the agent only.

Does Using Devql access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Using Devql safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Using Devql use?

Using Devql is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Using Devql use?

About 1.6k tokens (SKILL.md is roughly 6.3k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Using Devql?

Skills that share tags, products or a category with Using Devql: Ask Sonner (leepokai/Codync, 177 stars), Mobile Native (leepokai/Codync, 177 stars), Auto Improve (crimeacs/auto-improve, 135 stars) and Run Shunt (pleaseai/shunt, 264 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Using Devql?

bitloops (a GitHub organization) maintains it in bitloops/bitloops, which has 243 GitHub stars. The repository holds 2 skills in this directory. The repository was last updated on June 1, 2026.

Source: bitloops/bitloops on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.