reference-hunt

reference-hunt is a skill for Claude Code from GreatMark/fable-field-guide-skills. It costs 170 tokens per session (1,345 once invoked), scanned A, original, MIT.

A method for studying how an existing product, library, or codebase works by reading its source code and tests.

In plain words
What is it for?
Use it to find a reference implementation, inspect its edge cases and error handling, and adapt its logic to another project.
Why use it?
It provides concrete implementation details when screenshots or documentation do not explain the behavior you need to reproduce.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the fable-field-guide-skills plugin — 9 skills, 2 hooks shipped together

Good fit Use it to find a reference implementation, inspect its edge cases and…

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/greatmark/fable-field-guide-skills/reference-hunt
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.

Any agent
npx skills add GreatMark/fable-field-guide-skills --skill reference-hunt
Clone the repo
git clone --depth 1 https://github.com/GreatMark/fable-field-guide-skills

Made for: Claude Code.

Or install fable-field-guide-skills, the plugin that ships this one along with the rest of its 9 skills, 2 hooks.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for reference-hunt

README.md
[![agentmods](https://agentmods.dev/badge/skills/greatmark/fable-field-guide-skills/reference-hunt.svg)](https://agentmods.dev/skills/greatmark/fable-field-guide-skills/reference-hunt)
Your own site
<a href="https://agentmods.dev/skills/greatmark/fable-field-guide-skills/reference-hunt"><img src="https://agentmods.dev/badge/skills/greatmark/fable-field-guide-skills/reference-hunt.svg" alt="Measured on agentmods" height="20"></a>
Per session 170 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,345 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00170 $0.01345
Opus 5 $0.00085 $0.00673
Sonnet 5 $0.00034 $0.00269
Haiku 4.5 $0.00017 $0.00135

Measured 7d ago against content hash f2bf91a015a1, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

reference-hunt scanned grade A with 0 findings 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 7d 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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/reference-hunt/SKILL.md · 98 lines

How it starts

The opening of the file, as written. The whole thing — 98 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Reference Hunt

Sometimes the user cannot describe what they want in words — showing beats telling. The best reference is source code.

Workflow

  1. Get a pointer. If the user already named a reference (a library, a vendored crate, a folder, a component on a website), start there. If not, propose 2-3 candidate references yourself — from the codebase, vendored dependencies, or well-known open-source implementations — and let the user pick. One short message, not an interview.
  2. Read the real implementation. Locate the source in this order: node_modules/vendored copies inside the project → the package's repository at the installed version (npm view <pkg> repository or the ecosystem's equivalent, then fetch the tagged file) → GitHub code search (needs auth; anonymous fallback: shallow-clone the repo or fetch raw files by path from raw.githubusercontent.com). For a website module, read the underlying markup and code. Read the source, not its README, docs, or a screenshot. Also skim its test file — edge-case semantics (ordering, error paths, boundary values) live in tests more reliably than in the implementation body. A different language from the target is fine; semantics transfer, syntax does not. When the reference's framework differs from the target's, list which behaviors are carried by framework scheduling (pre-paint effects, batched updates, subscription timing) — re-create those deliberately; they neither transfer for free nor count as mere syntax.
  3. Write down what you are taking. Before implementing, give the user a short list, in the language of the conversation: the semantics being ported and what is deliberately not being ported. Cover five dimensions: inputs/API shape, outputs and return values, state transitions, edge cases, failure behavior. The list must outlive the chat:
    • if implementation starts now, record the ported semantics under Context and your own deliberate deviations under Decisions in implementation-notes.md (per the implementation-notes skill);
    • otherwise write it to porting-checklist.md in the project root, headed by the reference name+version, then the five dimensions; if planning comes next, hand that file to implementation-plan as an input. Do not commit it unless the user asks; in a git repo, exclude it locally: n=porting-checklist.md; p="$(git rev-parse --git-path info/exclude 2>/dev/null)" && { mkdir -p "$(dirname "$p")" && { grep -qxF "$n" "$p" || echo "$n" >> "$p"; }; } 2>/dev/null || echo "note: could not exclude — leaving it untracked, never staging it"
  4. Reimplement, don't transplant. Port the behavior into the target stack following the project's existing conventions and idioms. Cite the reference file paths — with version or commit for remote references (e.g. [email protected] debounce.js) — in your summary so claims are checkable.

Read the full file on GitHub · 98 lines

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. 7d ago First seen · 98 lines · 170 tokens per session scan A f2bf91a015a1

Subscribe to this mod's changes

reference-hunt is a skill published in the GitHub repository GreatMark/fable-field-guide-skills (9 stars, last pushed 29d ago), licensed MIT. It adds 170 tokens to every session and 1,345 once invoked, about $0.0009 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

idea-refine

Refines raw ideas into sharp, actionable concepts through structured divergent and convergent thinking. Use when an idea is still vague, when you need to stress-test assumptions before committing to a plan, or when you want to expand options before converging on one. Triggers on "ideate", "refine this idea", or…

addyosmani/agent-skills · 75 tokens

chinese-documentation

A Chinese technical-documentation style guide covering spacing, punctuation, numbers, terminology, and links when Chinese and English appear together.

jnMetaCode/superpowers-zh · 62 tokens

chinese-git-workflow

A reference for configuring Git with Chinese code-hosting services such as Gitee, Coding.net, GitLab China, and CNB, including SSH, HTTPS, credentials, CI, and repository mirroring.

jnMetaCode/superpowers-zh · 69 tokens

skillshare-implement-feature

Implement a feature from a spec file or description using TDD workflow. Use this skill whenever the user asks to: add a new CLI command, implement a feature from a spec, build new functionality, add a flag, create a new internal package, or write Go code for skillshare. This skill enforces test-first development…

runkids/skillshare · 114 tokens

setup-matt-pocock-skills

A setup skill that configures engineering skills for a repository, including its issue tracker, labels, and documentation layout. A repository is the project folder managed by version control.

devcxl/mattpocock-skills-zh · 43 tokens

create-slack-app

Use when a developer wants to create, scaffold, or bootstrap a new Slack app or agent from scratch with the Slack CLI. Covers prerequisites, sandbox setup, authentication, and creating + running a project from a Bolt (JS or Python) template locally. Trigger on "create a Slack app", "new Bolt app", "start a Slack…

slackapi/slack-skills-plugin · 74 tokens