skills/plan/SKILL.md
Produce a revision plan for the surrounding text based on the prompt. Supports multi-turn conversation chains.
npx skillsauth add a554b554/Reactant planInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Analyze the surrounding text and return a structured revision plan. Supports multi-turn conversation where the user can refine the plan iteratively before resolving.
<@plan: ...> tag appears, including any prior <@plan>/<@output> pairs in the chain.<@plan>/<@output> pairs in the tag chain. If they exist, treat them as conversation history.<@output: ...> tag immediately after the current <@plan> tag. The format must be exactly: <@output: your plan here> do NOT use a closing </@output> tag, and do NOT place content outside the tag. Keep the original content and all <@plan>/<@output> tags intact.After receiving an <@output>, the user may:
<@output> directly — modify the plan text, then use <@resolve> to apply it.<@plan:> tag after the <@output> — this continues the conversation. The new <@plan> prompt is treated as a follow-up that can refine, extend, or redirect the previous plan.Each <@plan>/<@output> pair forms one turn. When producing a new <@output>, read the full chain of prior turns as conversation history. All prior pairs stay intact — only append a new <@output> after the latest <@plan>. The output plan should always be included in the <@output> tag, e.g., <@output: your plan here>. The conversation chain is terminated by <@resolve>, which consumes the entire chain.
The surrounding text may contain scope modifier tags. When producing a plan, take these into account:
<< >>: Text inside << and >> is locked. Do not suggest changes to protected regions in your plan.(( )): Only text inside (( and )) is editable. Limit your suggestions to field regions only.Preserve all scope modifier delimiters as-is in the output — they will be consumed later by <@resolve>.
The prompt may contain context references wrapped in double backticks ( ``). These point to external resources (file paths, section titles, etc.) that you should look up.
<@plan: how to align this with ``intro.md``> — read intro.md and incorporate its content into your analysis.<@output: the plan is written in ``path``> after the <@plan> tag so <@resolve> knows where to find it. delimiters in the output (they will be consumed later by <@resolve>).data-ai
Apply a preceding plan/output chain to the content, then remove all tags.
tools
Proofread surrounding text for grammar and spelling errors.
data-ai
Generate and execute a visualization script for a given data file based on user instructions and context.
content-media
Fill a placeholder — infer the best content from surrounding context.