npx agentspec-cli install 0d698391-76ed-474a-aca6-d659b4f327d8 --framework ClaudeCode→ .claude/skills/ph/SKILL.md
# Placeholder ## Activation Boundary This is a tag-only worker skill. Use it only after `execute` or `execute-one` selects an exact `<@ph...>` annotation in a target document and hands off its context. Never activate it from ordinary chat, semantic similarity, or any skill other than those two routers. Without that handoff, produce no placeholder and direct the user to add the tag and run a router. Generate content that fits naturally at the tag position and replace only the tag with it. ## Tag ``` <@ph> <@ph: hint about what to fill> ``` Example: `We evaluated <@ph: the number of participants> participants.` ## Input - **File path** — the document being processed. - **Surrounding block** — the maximal run of non-blank lines containing the tag; blank lines bound it. - **Prompt** (optional) — a hint (for example "a transition sentence"). Default: infer from context. ## Workflow 1. Read the text before and after the tag; read the full file if broader context helps. 2. Generate content that fits naturally at that exact position in tone, style, and meaning. 3. Write the result back: replace ONLY the `<@ph>` tag with the generated text. All surrounding text stays untouched. ## Scope Modifiers `<@ph>` never modifies surrounding content, so modifiers mainly provide context: - **Protect `<< >>`** — context only; never modify. This skill does not remove protect delimiters outside its own insertion. - **Field `(( ))`** — when the tag sits inside a field, generate content that belongs in that field. ## Context References Double-backtick references in the prompt point to resources that inform the generated content (for example `<@ph: a summary based on ``results.md``>`). Resolve paths relative to the document's directory unless absolute. ## Rules - The output must read as if it had always been part of the text; keep it concise and contextually appropriate. - Replace only the selected annotation. Preserve every unrelated annotation, `<@comment: ...>` tag, `<@chat-model>` block, and contextual literal/example byte-for-byte. - On any error (for example an unreadable referenced file), fail cleanly: leave the tag, change nothing, report.
Built on the Agent Skills standard
paper-context-resolver
Optional narrow helper skill for README-first AI repo reproduction. Use only when the README and repository files leave a narrow reproduction-critical gap and the task is to resolve a specific paper detail such as dataset split, preprocessing, evaluation protocol, checkpoint mapping, or runtime assumption from primary paper sources while recording conflicts. Do not use for general paper summary, repo scanning, environment setup, command execution, title-only paper lookup, or replacing README guidance by def
by AgentHub Crawler
🚨 CRITICAL: Relay-First Communication Rule
**When you receive a relay message from another agent (marked `Relay message from [name]`), you MUST respond ONLY via relay protocol. NEVER respond with direct text output.**
by AgentHub Crawler
QuizApp Repository Guidelines
Use these skills for detailed patterns on-demand. Always read the SKILL before writing code that falls under its trigger.
by AgentHub Crawler
Sign in to leave a comment.