Cowork session planner
Skill KirKruglov/claude-skills-kit/skills/sessions-and-context/cowork-session-planner
70+ curated agent skills for Claude Cowork and Claude.ai — ready-to-use tools for non-technical users: project management, productivity, and AI workflow automation
npx -y skills add KirKruglov/claude-skills-kit --skill cowork-session-plannerAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 14 stars14 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Generate a pre-session brief for Cowork: current status, session goal, and prioritised work plan from local project files — before your first real message. Use when starting a session, returning after a break, or needing focus. Triggers: 'session brief', 'prepare my session', 'start my cowork session', 'бриф сессии', 'подготовь сессию', 'что делать в этой сессии'.
SKILL.md
5.3 KB, ~1.2k tokens by cl100k_base, as published. Nobody here has run it
Cowork Session Planner
This skill generates a pre-session brief for non-technical Cowork users by scanning available project files (context.md, logs, tasks, CLAUDE.md) and producing a minimal structured brief: current project status, recommended session goal, and a prioritised work plan. It runs at the very start of a session — before any real work task is requested — so the user knows exactly where to focus.
Input: Current workspace / project folder. Optional: explicit session goal as free text.
Output: Structured markdown brief (inline response). Optionally saved to output/session-brief.md on user request.
Language Detection
Detect the user's language from their message:
- If Russian (or contains Cyrillic): respond in Russian
- If English (or other Latin-script language): respond in English
- If ambiguous: respond in the language of the trigger phrase used
Instructions
Step 1: Parse Input
-
Check if the user provided an explicit session goal (free text after the trigger phrase).
- If yes: store as
stated_goal; use it as the primary session focus in Step 4. - If no:
stated_goalis empty; derive focus from project files in Step 3.
- If yes: store as
-
Identify the project folder to scan.
- Default: current workspace / selected folder.
- If user specifies a path, use that path.
Step 2: Scan Project Files
-
Look for the following files in the workspace (in order of priority):
context.mdorCLAUDE.md— project overview and statuslogs/log.md— recent activity logTASKS.mdor any file containing open tasks- Any
output/skill-spec.mdor in-progress output files
-
For each found file, extract:
- context.md / CLAUDE.md: project name, current status, open questions
- logs/log.md: last log entry (date + action description)
- TASKS.md: open tasks (status: pending or in_progress)
- skill-spec.md: if present, note "skill spec in progress"
-
Note any missing files for the Notes section of the brief.
Edge Cases:
- If no context.md or CLAUDE.md found: proceed without project overview; flag as missing in Notes.
- If log file is empty or absent: skip last-action extraction; open brief with "No activity log found."
- If multiple task files found with conflicting state: list all sources; flag conflict in Notes.
Step 3: Determine Session Focus
-
If
stated_goalis set: use it as the session goal. Skip to Step 4. -
Otherwise, derive focus from extracted data:
- Check open tasks (status: pending / in_progress) — select the highest-priority item.
- If no tasks, check last log entry — infer next logical step from last completed action.
- If neither available: propose "Define session goal" as the first task.
-
Compile a prioritised work list (max 5 items):
- Item 1: derived or stated session goal
- Items 2–5: remaining open tasks or logical next steps
Step 4: Assemble Brief
-
Build the pre-session brief with these sections:
- Project / Status: project name + 1-sentence current state
- Session Goal: recommended or stated goal (1 sentence)
- Work Plan: numbered list, max 5 items
- Blockers: any open blockers or conflicts from Step 2; "None" if clean
- Notes: any missing files, file conflicts, or flags from Step 2
-
Use bilingual headers (EN + RU) if Russian was detected; English-only otherwise.
-
Display the brief inline. Do not save to file unless user explicitly requests it.
Step 5: Handle Save Request
- If user says "save" or "save to file": write brief to
output/session-brief.md. - Confirm file written and provide path.
Negative Cases
- No project files found (empty workspace): Return: "No project files found. Please select your project folder or provide a session goal to get started." Do not generate an empty or partial brief.
- User invokes mid-session (active tasks detected): Detect in-progress tasks or recent log entries (within last hour). Note: "It looks like you're already mid-session." Offer a re-focus brief instead of an initial brief (same format, but titled "Re-focus Brief").
Output Format
Inline markdown response:
## Сессионный бриф / Session Brief
**Проект / Project:** [project name or folder]
**Статус / Status:** [1-sentence current state]
### Фокус сессии / Session Goal
[Recommended or stated session goal — 1 sentence]
### План работы / Work Plan
1. [Task 1 — highest priority]
2. [Task 2]
3. [Task 3]
...
### Блокеры / Blockers
- [Blocker 1 if any, otherwise: None]
### Заметки / Notes
- [Missing files, conflicts, or flags — if any]
Field rules:
- Work Plan: 1–5 items; each item is a specific, actionable task (not a vague category)
- Blockers: specific obstacle or "None"
- Notes: only populated when something is flagged (missing file, conflict, mid-session detection)
What ships with it: 4 files
24.6 KB alongside SKILL.md
docs/
- USER-GUIDE.md6.0 KB
- USER-GUIDE.ru.md9.7 KB
- README.md3.5 KB
- README.ru.md5.4 KB
Gives 0 of the 12 instructions most plan spec skills give in ~1.2k tokens
Counted across 1,099 of the 1,860 authors here whose files we hold, read 2026-08-07
- Ask one question at a timein 51 of 1099
- Break plans into vertical slicesin 29 of 1099, across 11 files
- Publish issues in dependency orderin 27 of 1099, across 9 files
- Iterate until user approves the breakdownin 25 of 1099, across 7 files
- Explore the repository to understand the codebase statein 24 of 1099, across 7 files
- Use domain glossary vocabularyin 23 of 1099, across 5 files
- Apply correct triage labels to published issuesin 23 of 1099, across 5 files
- Prefer AFK slices over HITLin 22 of 1099, across 7 files
- Write a specification before writing any codein 22 of 1099, across 14 files
- Write failing tests before implementation codein 22 of 1099, across 20 files
- Ask clarifying questions until requirements are concretein 21 of 1099, across 13 files
- Respect existing architecture decision recordsin 20 of 1099, across 5 files
Said here and by no other author read
- respond in the detected user language
- check for an explicitly stated session goal
- identify the target project folder
- scan project files for context and status
- extract open tasks from project files
- note any missing or conflicting files
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.