跳到正文
cccskills
免费在 GitHub 公开

activity-recap

Use when the user needs a standup-ready summary of recent project activity from local evidence.

中文简介正在准备中。当前显示原文说明。

查看安装方法

包含的文件(1)

  • SKILL.md3.6 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

activity-recap

Lifecycle stage: LEARN

Trigger

Use when the user asks what changed recently, what they worked on, what a team did, or needs a short standup/status recap.

Do not use when the user asks for a product roadmap, a release note, or a narrative post. Those need planning, review, or writing skills instead.

When not to use

Do not use when this trigger is absent; choose the command or skill that owns the requested state, artifact, and verification gate.

Inputs

  • Repository root or parent workspace.
  • Date range or working-day window.
  • Optional author filter.
  • Optional branch or remote scope.
  • Desired output format: bullets, grouped-by-repo, grouped-by-author, or handoff note.
  • Privacy constraints: files, authors, branches, or messages to omit.

Procedure

  1. Confirm or infer the workspace root.
  2. Discover repositories within the requested depth without crossing unrelated private directories.
  3. Inspect local history for the requested date range and filters.
  4. Group activity by repository, author, and theme.
  5. Normalize commit messages into human-readable bullets without inventing intent.
  6. Include file-change stats only when they clarify the work.
  7. Mark repositories with no matching activity as empty only when the user asked for a full workspace scan.
  8. Produce a concise recap with the inspected range and evidence summary.

Anti-Rationalization

ShortcutRebuttal
"I remember what changed."Recaps must come from checked local history, files, or logs, not memory.
"A high-level summary is enough."Name the inspected scope and timeframe so readers can trust the boundary.
"No activity means nothing happened."Say no matching committed activity was found only for the scope actually checked.

Verification

  • State the date range and filters used.
  • State which repositories were inspected.
  • Name the recap artifact target and the command, log, or file evidence used to produce it.
  • Cite commit hashes, counts, or changed-file summaries when useful.
  • If no activity is found, say so and list the checked scope.
  • Do not infer uncommitted work unless separately inspected and explicitly labeled.

Output Artifact

Produce a concise recap that names the inspected scope, timeframe, filters, evidence checked, notable changes, empty scopes when relevant, assumptions, blockers, and the next action.

Failure Modes

  • Wrong author identity: ask for or list available authors instead of guessing.
  • Shallow or stale repository: label the result as local-only or fetch only if allowed.
  • Bad commit messages: summarize conservatively from file paths and stats.
  • Private branch content: respect privacy constraints and omit sensitive details.
  • Multi-repo noise: group by repository and suppress empty projects unless requested.

Example

Trigger: user needs a recap of recent project activity. Action: inspect local commits, files, branches, and relevant logs before summarizing. Output artifact: activity recap with checked scope, blockers, and next action. Verification: cite command output, author/date filters, repositories checked, and unknowns.

Activity recap for 2026-05-13, author filter: current Git user.

- agent-brain: hardened repository validation and added CI quality gates.
  Evidence: 5 commits inspected, tests referenced in commit messages.
- website: no matching committed activity in the inspected range.

Checked: 2 repositories under the workspace root.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Use when a command, adapter, or runtime smoke depends on proving an agent runtime's capabilities before trusting command routing, writes, shell access, or validation claims.

中文简介正在准备中。当前显示原文说明。

rohitg00/agentbrain412026年5月27日 更新

Use when agent-produced work needs a safety, evidence, and reliability check before handoff, merge, or trust.

中文简介正在准备中。当前显示原文说明。

rohitg00/agentbrain412026年5月27日 更新

Use when an output artifact, template, schema, example, or handoff field must stay aligned across commands, skills, docs, and validators.

中文简介正在准备中。当前显示原文说明。

rohitg00/agentbrain412026年5月27日 更新

Agent Brain /brain-brief: after intake, research, and grill have enough signal.

中文简介正在准备中。当前显示原文说明。

rohitg00/agentbrain412026年5月27日 更新

Agent Brain /brain-build: an implementation plan has a selected task and validation method.

中文简介正在准备中。当前显示原文说明。

rohitg00/agentbrain412026年5月27日 更新

Agent Brain /brain-design: a product brief needs ux or interaction design before planning.

中文简介正在准备中。当前显示原文说明。

rohitg00/agentbrain412026年5月27日 更新

查看 rohitg00 的全部技能

报告此技能的问题