Install
$ agentstack add skill-t0ddharris-claude-code-skills-brief ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ● Environment & secrets Used
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Session Brief
Run this at the end of a session to capture context and checkpoint all work to GitHub.
Steps
1. Review Existing Brief
Read brief/session-brief.md to understand what was there from last session.
2. Write the New Brief
Overwrite brief/session-brief.md with a comprehensive brief covering:
- Date, time, and session number at the top — run
date "+%A %Y-%m-%d %H:%M %Z"to get the real date, day of week, and local timezone rather than guessing (LLMs are unreliable at computing the day of the week, especially for future dates) - Important progress made in the current session
- Key decisions and architectural changes
- Unfinished tasks and next steps
- Technical details that need to be preserved
- Any context that would be lost between sessions
This is NOT about brevity — be as thorough as needed to retain all important context.
2.5. Reflect on Skills
After writing the brief, run the reflect skill to scan the conversation for learnings. This is a delegation, not a repeat of the reflect instructions. Load and follow the reflect skill.
- If learnings are found: present them to the user alongside the brief, before asking about git
- If no learnings: report "no new skill learnings detected" in one line and continue to Step 3
- Any approved skill updates will be committed together with the brief in Step 4
2.7. Sync Skills (opt-in projects only)
If the project has an .agents/ directory, run the sync-skills skill with project scope (no prompt) so .agents/skills/ mirrors .claude/skills/ before committing. Only mention it if something was added or removed. If there is no .agents/ directory, skip silently — never create one.
3. Ask About Git
After writing the brief, always ask the user which option they want:
- Commit + push — full checkpoint to GitHub
- Commit only — local save point, push later
- Brief only — just the file update, no git operations
Present these as a quick choice. Do NOT assume commit+push.
4. If Committing (options 1 or 2)
- Run
git statusto see all changes (staged + unstaged + untracked) - Run
git log --oneline -3to match commit message style - Stage all relevant files (
git add— use specific paths, not-A) - Commit with message format:
Session [N]: [1-2 sentence summary of session work] - If option 1, push to origin:
git push
Commit rules:
- Do NOT stage files that look like secrets (
.env, credentials, tokens) - If there are uncommitted changes unrelated to this session, ask the user before including them
- If push fails (e.g., remote is ahead), stop and ask the user — do not force push
When to Use
- At the end of significant sessions
- When important decisions have been made
- Before ending a session where critical details might be lost
- When the user explicitly requests
/brief
Read-Only Mode
When asked to "read the brief" (typically at session start via /start), just read the file without modifying it. Do not commit or push in read-only mode.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: t0ddharris
- Source: t0ddharris/claude-code-skills
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.