project skill
✓ VerifiedCreate and manage project thinking notebooks in Obsidian to track goals, decisions, open questions, and lessons learned.
- GitHub stars
- 16
- Files
- 1
- Last commit
- Oct 28, 2026
- License
- MIT
- Allowed tools
- —
What does the project skill do?
project creates and manage project thinking notebooks in Obsidian to track goals, decisions, open questions, and lessons learned.
Summary generated from the skill's SKILL.md.
About the project skill
Managing project thinking notebooks within an Obsidian vault is what this skill enables for users tracking finite efforts. It allows an agent to create, update, or archive dedicated notes that capture goals, decisions, open questions, and lessons learned. When an agent identifies a need for a project record, it searches the vault for existing templates and validates that the work is a finite project rather than an ongoing area.
Users should load this skill when they want to document a specific endeavor without duplicating task tracking systems. The tool relies on the vault safe-write flow to ensure data integrity and interacts with files like SKILL.md and AGENTS.md to follow local schemas. It produces structured project files, metadata, and updated index links. The skill distinguishes itself by focusing on high-level thinking and decision logs rather than managing individual tasks, which should remain in their original tracking systems.
What an agent can do with it
- 1Start a new project notebookCreate a structured thinking record for a new finite effort by following the active project template.
- 2Document project decisions and lessonsCapture new decisions, questions, or lessons learned by updating an existing canonical project note.
- 3Archive a finished projectPerform a retrospective and move a completed project note to its designated archive destination.
- 4Distinguish projects from ongoing areasVerify if a piece of work is a finite project with a checkable outcome or an ongoing standard.
Written from the skill's SKILL.md. Every use case is checked against a quote from it.
How to install the project skill
Each command copies the skill folder into the client's skills directory. Restart the client or start a new session to load it.Install the project skill in Claude Code
~/.claude/skills/project/SKILL.md
$ npx degit kriscard/Skills/skills/obsidian/project ~/.claude/skills/projectInstall the project skill in Claude Code (project)
.claude/skills/project/SKILL.md
$ npx degit kriscard/Skills/skills/obsidian/project .claude/skills/projectInstall the project skill in Codex
~/.agents/skills/project/SKILL.md
$ npx degit kriscard/Skills/skills/obsidian/project ~/.agents/skills/projectInstall the project skill in Gemini CLI
~/.gemini/skills/project/SKILL.md
$ npx degit kriscard/Skills/skills/obsidian/project ~/.gemini/skills/projectInstall the project skill in Claude.ai
npx degit kriscard/Skills/skills/obsidian/project project && zip -r project.zip project # claude.ai → Customize → Skills → + → Create skill → Upload a skill → project.zip
Automated checks
Ran Nov 16, 2026Full report →Passed our automated checks. That is not a security audit: read the scripts before you let an agent run them. How checks work
- Source repository linked and reachablekriscard/skills/skills/obsidian/project is public and reachablegithub.com/kriscard/skills/tree/HEAD/skills/obsidian/project ↗Pass
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit 2026-10-28github.com/kriscard/skills/commits ↗Pass
- No download-and-run or credential theft in scriptsNo scripts in the skill folderPass
- No hidden instructions in SKILL.mdNo override phrases, safeguard switches, credential requests or hidden charactersgithub.com/kriscard/skills/blob/HEAD/skills/obsidian/project ↗Pass
- Pre-approved tools are scopedNo pre-approved tools; the agent asks before using onegithub.com/kriscard/skills/blob/HEAD/skills/obsidian/project ↗Pass
- SKILL.md follows the Agent Skills formatValid name and description; says when to use it; 82 linesgithub.com/kriscard/skills/blob/HEAD/skills/obsidian/project ↗Pass
SKILL.md
View on GitHubProject
Keep a thinking record for a finite effort in the vault's live Projects location. Use it for goals, decisions, questions, and lessons; link to the task tracker instead of duplicating it. Apply deadline and metadata requirements only when the live schema defines them.
Choose a mode
- Create: start a project thinking notebook.
- Update: capture a decision, question, thought, or lesson.
- Complete: done, stopped, or archive.
Ask when the mode is ambiguous.
Create
1. Search and validate
Read ../vault/SKILL.md and AGENTS.md, then discover and read the active Project template.
Search the live Projects location and inspect close matches before proposing a duplicate.
Confirm the work is a Project rather than an Area: it must be an active finite effort with a checkable outcome. If it is an ongoing standard, propose an Area instead.
2. Understand the project
Ask no more than four concise questions in one structured batch when needed:
- What are you trying to finish?
- What have you decided, and why?
- What questions are still open?
- What thoughts or lessons are worth keeping?
Use answers already present in the conversation instead of asking again. Gather additional detail only when the schema requires it.
3. Choose structure and preview
Default to one project file. Use a project folder only when the work already needs durable supporting notes. Fill the live template with supported content; leave unanswered sections open rather than inventing progress or commitments. Show the path, metadata, and complete draft before writing.
After approval, create through the vault safe-write flow, then verify the page and any separately
approved project-index link.
Update
Resolve and read the canonical project note. Compare its recorded state with user-provided evidence. Draft only the sections that changed and distinguish observed staleness from assumptions.
Show the exact change, obtain approval, update safely, and reread. Do not rewrite supporting or
human-owned notes without permission from AGENTS.md and the user.
Complete
Confirm whether the outcome was achieved, deliberately stopped, or superseded. Offer a concise retrospective before archiving. Preview:
- final status and outcome;
- lessons or decisions worth retaining;
- the live schema-defined archive destination;
- affected project-index links.
Approval to record completion is separate from approval to archive. Write the approved final state; move the note only after the archive destination is explicitly confirmed. Update approved links and verify both source absence and destination presence when a move occurs. Reusable knowledge stays in Knowledge; project completion does not archive it.
Completion
Report the canonical path, mode, confirmed outcome/state, writes performed, index changes, and any unresolved schema or evidence gap.
Files in this skill
1project skill questions
How do I install the project skill in Claude Code?
Run npx degit kriscard/Skills/skills/obsidian/project ~/.claude/skills/project. Claude Code loads skills from ~/.claude/skills (all projects) or .claude/skills (one project) and uses a skill when its description matches the task.
Which clients can use the project skill?
Any client that reads the Agent Skills format: Claude Code, Codex, Gemini CLI and claude.ai (upload as a ZIP). Scripts in the skill run only where the client allows code execution.
Should I use this skill for task tracking?
No, keep task tracking in its existing system and only use this skill when a vault artifact is requested.
What is the difference between a Project and an Area in this workflow?
A Project must be an active finite effort with a checkable outcome, whereas an Area is used for ongoing standards.
What happens to knowledge when a project is completed?
Reusable knowledge stays in the Knowledge section; project completion does not archive it.