Skip to content
cccskills
FreePublished on GitHub

shared-content-references

Store instructional prose once and reference it from multiple skills or agents instead of duplicating it. Use when the same steps, rules, or reference material appear in two or more SKILL.md or agent files in one plugin, when deciding whether a shared doc belongs in the plugin-root docs directory or a single skill's references directory, or when tempted to use a symlink or a ../ path to share a file between components.

The English summary is being prepared. Showing the original description.

See how to install

Files included (4)

  • SKILL.md5.2 KB
  • references/annotation-format.md1.0 KB
  • references/verification.md891 B
  • references/what-not-to-use.md889 B

SKILL.md (original)

Review the instructions the agent will receive before you install.

Shared Content References

Resolve "where does the shared copy live?" for prose duplicated across two or more SKILL.md or agent files inside one plugin.

1. Scope Gate

Applies when the same prose is needed by 2+ SKILL.md or agent files in one plugin.

Does not apply — route elsewhere instead:

  • One skill's own content is too long → activate the /plugin-creator:refactor-skill skill (SK006/SK007 token thresholds).
  • Deciding whether content belongs in a skill's SKILL.md body or its own references/ directory (a single-skill boundary, not a cross-skill one) → activate the /plugin-creator:optimize skill or the /plugin-creator:agentskills skill.
  • Sharing executable code (scripts, shell utilities) rather than prose → activate the /plugin-creator:component-patterns skill, §Shared Resources (lib/ at the plugin root, accessed via ${CLAUDE_PLUGIN_ROOT}/lib/).

2. Placement Decision

flowchart TD
    Start(["Same prose needed in 2+ places"]) --> Q1{"Consumed by components<br>in 2+ skill directories,<br>or by an agent<br>(agents have no references/)?"}
    Q1 -->|"Yes"| Root["Place in plugin-root docs/<br>Reference via Technique 1"]
    Q1 -->|"No — one skill plus<br>the agents it dispatches"| SkillRef["Place in that skill's references/<br>Reference via the ${CLAUDE_SKILL_DIR} form<br>from Technique 2"]
    Root --> Q2{"More than ~3 shared docs<br>accumulate in this location?"}
    SkillRef --> Q2
    Q2 -->|"Yes"| Index["Add an index skill (Technique 2)<br>so consumers load one skill<br>instead of repeating paths"]
    Q2 -->|"No"| Direct["Reference the doc directly<br>from each consumer<br>No index skill needed yet"]

3. Technique 1 — Plugin-Root Shared Doc

[<doc name>](../../docs/<doc>.md) — contains <what>; read before <when>.

From skills/<name>/SKILL.md the relative link stays inside the plugin and needs no harness support, so it is the primary form for a doc consumed across 2+ skill directories. Claude Code also substitutes ${CLAUDE_PLUGIN_ROOT} in plain SKILL.md prose and markdown link targets; write that form only in a plugin that targets Claude Code alone. Load verification.md before relying on either form in a new, unverified runtime.

4. Technique 2 — The Index-Skill Pattern

An index skill's own SKILL.md lists each shared doc with the annotation contract from annotation-format.md:

[<doc name>](${CLAUDE_SKILL_DIR}/references/<doc>.md) — contains <what>; read before <when>.

Every consumer skill or agent writes one sentence and never repeats a path:

For <topic>, activate the /plugin-name:index-skill-name skill.

Moving or renaming a shared doc becomes a one-file edit inside the index skill.

Shared content lives inside the plugin that ships it. A doc in the surrounding repo is not shared content — an installed consumer never has that tree, so the reference resolves for the author alone.

5. Reference Files

Confirm each shared reference is present in every environment, bundled and reached by a relative path inside the plugin, or inlined; otherwise inline, bundle, guard, or delete it. A harness variable counts only where that harness substitutes it.

Load annotation-format.md before writing any reference link created by Technique 1 or 2 — defines what every reference must state and why.

Load what-not-to-use.md before reaching for a symlink, a ../ path, or copy-paste to share content across components — each fails for a documented reason.

Load verification.md after adding or changing any shared reference — canary, skilllint, and target-existence checks.

6. Related Skills

  • /plugin-creator:optimize — the SKILL.md-vs-references/ boundary within one skill; load first if the question is about one skill's own structure, not sharing across skills.
  • /plugin-creator:agentskills — the Agent Skills Open Standard for cross-agent-product portability. ${CLAUDE_PLUGIN_ROOT} and ${CLAUDE_SKILL_DIR} are Claude-Code-specific and do not carry over to a portable skill.
  • /plugin-creator:refactor-skill — one skill exceeding SK006/SK007 token thresholds; a different problem from N skills sharing content.
  • /plugin-creator:component-patterns — §Shared Resources covers shared executable code (lib/ at the plugin root), the code counterpart to this skill's shared prose.
  • The plugin-assessor agent's cross-component duplicate-content check routes here for the remedy.
  • The start-refactor-task skill's CREATE shared references if specified step routes here for what a shared reference is and how to create one.

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

Add automated documentation updater to any Claude skill. Creates a Python sync script that downloads upstream docs, processes markdown for AI consumption, and maintains local cache with configurable refresh. Collects template variables, then delegates implementation through 5-phase workflow. Use when adding auto-updating reference documentation to plugins or skills.

The English summary is being prepared. Showing the original description.

Jamie-BitFlight/claude_skills67Updated Oct 20, 2026

SAM-style feature initiation workflow — discovery through codebase analysis, architecture spec, task decomposition, validation, and context manifest. Use when a user asks to add a feature, plan a feature, or convert an idea into an executable SAM plan.

The English summary is being prepared. Showing the original description.

Jamie-BitFlight/claude_skills67Updated Oct 20, 2026

Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction.

The English summary is being prepared. Showing the original description.

Jamie-BitFlight/claude_skills67Updated Oct 20, 2026

Browser automation for AI agents using the agent-browser CLI and Playwright. Use when navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, logging into sites, or automating any browser task. Triggers on "open a website", "fill out a form", "click a button", "scrape data", "test this web app", "automate browser actions", or any programmatic web interaction request.

The English summary is being prepared. Showing the original description.

Jamie-BitFlight/claude_skills67Updated Oct 20, 2026

Runs the description-drift experiment — spawns all Claude Code agents simultaneously to collect self-reported capabilities, then compares them against static frontmatter descriptions to reveal how reliable orchestrator routing based on descriptions actually is. Use when measuring description drift across the agent fleet, re-running the capability collection experiment, analyzing a specific agent's self-reported capabilities, or auditing whether frontmatter descriptions accurately reflect agent behavior.

The English summary is being prepared. Showing the original description.

Jamie-BitFlight/claude_skills67Updated Oct 20, 2026

Create or adapt Claude Code agent definitions. Use when creating an agent, changing subagent configuration, selecting agent scope, or designing a specialized delegation role.

The English summary is being prepared. Showing the original description.

Jamie-BitFlight/claude_skills67Updated Oct 20, 2026

See all skills by Jamie-BitFlight

Report a problem with this skill