adev:research

A structured investigation of a technical topic using the project's code, web sources, and optionally GitHub repositories, producing a research document.

In plain words
What is it for?
Use it to investigate an approach, compare options, learn how other projects solve a problem, or collect best practices. It can link the findings to an existing issue.
Why use it?
It gathers evidence and organizes findings so decisions are based on more than guesswork or scattered searches.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/agentic-development/adev-plugin/research
Any agent
npx skills add agentic-development/adev-plugin --skill research
Clone the repo
git clone --depth 1 https://github.com/agentic-development/adev-plugin

Made for: Claude Code, Codex.

Per session 87 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,393 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00087 $0.03393
Opus 5 $0.00044 $0.01697
Sonnet 5 $0.00017 $0.00679
Haiku 4.5 $0.00009 $0.00339

Measured 2d ago against content hash 0348b3732446, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

adev:research scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Instruction-override phrasingmediumPrompt injection

Text telling the model to disregard its earlier instructions or safety rules is the shape of a prompt injection, whoever wrote it.

- Phrase list: "ignore previous instructions", "from now on", "you are now", "instead of", "do not mention", "your new task", "as an AI"

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

providers/codex/skills/research/SKILL.md · 248 lines

How it starts

The opening of the file, as written. The whole thing — 248 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Research a Topic

Conduct structured research across multiple sources and produce an organized research artifact at .context-index/research/<slug>.md. Sources include the local codebase, web search, and GitHub repositories. Findings are synthesized into a single document with attribution, code examples, and actionable recommendations.

Announce at start: "I'm using the adev:research skill to investigate this topic."

Arguments

  • <topic> (required): free-text research topic (e.g., "dependency injection patterns in Node.js ESM")
  • --web: explicitly include web search as a source
  • --github <owner/repo>: include GitHub code search against the specified repository (must match owner/repo pattern)
  • --internal: explicitly include local codebase search as a source
  • --compare: organize findings as a comparison matrix with pros/cons/tradeoffs
  • --issue <id>: link this research to an existing issue on the issue board

Default Source Behavior

When no source flags are specified, the defaults are:

  • Web: enabled (if WebSearch tool is available in the researcher subagent's context)
  • Internal: enabled
  • GitHub: disabled (only used when --github <owner/repo> is explicitly provided)

When any source flag is specified, only the explicitly requested sources are used. For example, --internal alone means only the local codebase is searched.

Prerequisites

This skill requires .context-index/ to exist. If it does not, tell the user:

This project has not been initialized with the Agentic Development Framework. Run /adev:init first to set up the context index, then come back to research.

Do not proceed without a constitution.


Process

Complete these steps in order.

Step 1: Parse Arguments and Validate

  1. Extract the <topic> from the user's message. If no clear topic is provided, ask the user to clarify.
  2. Determine which sources to use based on flags (see Default Source Behavior above).
  3. If --github is specified, validate the value matches the owner/repo pattern (e.g., anthropics/claude-code). If invalid, print a warning and skip GitHub source.
  4. Verify .context-index/ exists. If missing, print the prerequisite message and stop.

Read the full file on GitHub · 248 lines

Files

What ships with it

5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 248 lines · 87 tokens per session scan B 0348b3732446

Subscribe to this mod's changes

adev:research is a skill published in the GitHub repository agentic-development/adev-plugin (11 stars, last pushed 3d ago), licensed MIT. It adds 87 tokens to every session and 3,393 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it B with 1 finding (instruction-override phrasing). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openai/codex · 114 tokens

imagegen

Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…

openai/codex · 113 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens