Help us improve
Share bugs, ideas, or general feedback.
From caspar
Use when the user wants to search the project's captured knowledge, recall a specific past learning, or discover what learnings already exist (e.g. "/caspar:recall", "what do we know about X", "is there a learning for Y"). Do NOT trigger to capture or write new knowledge — that is /caspar:learn.
npx claudepluginhub codename-inc/caspar --plugin casparHow this skill is triggered — by the user, by Claude, or both
Slash command
/caspar:caspar-recallThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Find and load the project's captured caspar learnings relevant to a query, so the answer is grounded in prior knowledge instead of re-derived.
Guides technical evaluation of code review feedback: read fully, restate for understanding, verify against codebase, respond with reasoning or pushback before implementing.
Share bugs, ideas, or general feedback.
Find and load the project's captured caspar learnings relevant to a query, so the answer is grounded in prior knowledge instead of re-derived.
$ARGUMENTS — the search query. Empty = list everything by category.{{project_root}}/.claude/skills/caspar-recall/SKILL.md (Claude Code), or{{project_root}}/.agents/skills/caspar-recall/SKILL.md (Codex).
Each registry row is skill-name|category|triggers|description. Live trigger keywords also sit in each loaded skill's frontmatter description.Read-only: the registry file above and the frontmatter descriptions already in context. No writes.
A grounded answer, plus the matching knowledge loaded into context. DONE when:
Skill({skill-name}) before answering.triggers + description and the trigger keywords in loaded skills' frontmatter.Skill({skill-name})./caspar:learn to capture it./caspar:learn once the current work yields something worth preserving. Do not fabricate entries.Return the loaded knowledge and a 1–2 line note on what was applied, in-thread. Persist nothing.
/caspar:learn.