capture
/capture
Store the current conversation transcript, agent output, or a git commit in VibeHub, and attach conversations to commits. Use when the user asks to save/capture/upload this session, transcript, or agent output to VibeHub, attach a conversation to a commit, or list/create VibeHub projects.
- Author
- aw3-technology ↗
- Repository
- vibehub-plugins ↗
- Licence
- MIT
- Last repository change
- 12.11.2026
Assessment and security
- Files in this package
- 1
- Scripts
- 0
- HTML comments
- 0
- URLs in SKILL.md
- 6
No signals detected. Repository changed on 12.11.2026, without references to retired models or APIs.
MIT, open: use and modify while retaining author attribution.
What it does and when to use it
Store the current conversation transcript, agent output, or a git commit in VibeHub, and attach conversations to commits. Use when the user asks to save/capture/upload this session, transcript, or agent output to VibeHub, attach a conversation to a commit, or list/create VibeHub projects.
Original author description
Store the current conversation transcript, agent output, or a git commit in VibeHub, and attach conversations to commits. Use when the user asks to save/capture/upload this session, transcript, or agent output to VibeHub, attach a conversation to a commit, or list/create VibeHub projects.
How it works
- Setup check
- Find this session's VibeHub id
- Attach this conversation to a commit
- One-shot: store a commit + transcript together
- Store agent output without the hooks
- Projects
- Reporting back
SKILL.md preview
---
name: capture
description: Store the current conversation transcript, agent output, or a git commit in VibeHub, and attach conversations to commits. Use when the user asks to save/capture/upload this session, transcript, or agent output to VibeHub, attach a conversation to a commit, or list/create VibeHub projects.
---
# VibeHub capture
VibeHub stores LLM conversation transcripts alongside the git commits they
produced. This skill is the manual capture path; the plugin's hooks already
stream every turn automatically when `VIBEHUB_TOKEN` is set.
## Setup check
All calls need two environment variables:
- `VIBEHUB_TOKEN` — bearer token (`vbh_…`), generated in the dashboard under
**Settings → API tokens**.
- `VIBEHUB_API_URL` — optional, defaults to `https://www.vibehub.co`.
Verify before doing anything else:
```bash
curl -sf -H "Authorization: Bearer $VIBEHUB_TOKEN" "${VIBEHUB_API_URL:-https://www.vibehub.co}/api/sdk/ping"
```
If this fails, tell the user to generate a token in the VibeHub dashboard
(Settings → API tokens) and export it as `VIBEHUB_TOKEN`, then stop.
## Find this session's VibeHub id
The hooks register every Claude Code session with
`externalId = claude-code:<session_id>`. The local map lives at
`~/.vibehub/claude-code/sessions.json` keyed by Claude session id. Read it to
find the VibeHub session id for the current session. If the file or entry is
missing, create a session first (see "Store output without the hooks").
## Attach this conversation to a commit
Use after committing, so the transcript shows up on the commit page:
```bash
curl -sf -X POST -H "Authorization: Bearer $VIBEHUB_TOKEN" -H "Content-Type: application/json" \
"${VIBEHUB_API_URL:-https://www.vibehub.co}/api/sdk/sessions/<SESSION_ID>/attach" \
-d "{\"sha\": \"$(git rev-parse HEAD)\", \"message\": $(git log -1 --pretty=%s | jq -Rs .), \"branch\": \"$(git branch --show-current)\"}"
```
Pass `projectId` too when known. Attaching closes the session; later turns in
the same Claude session will be captured into a fresh one by the hooks.
## One-shot: store a commit + transcript together
When there is no live session (or the user asks to capture a specific commit),
post commit metadata, diff, and conversation in one call:
```bash
curl -sf -X POST -H "Authorization: Bearer $VIBEHUB_TOKEN" -H "Content-Type: application/json" \
"${VIBEHUB_API_URL:-https://www.vibehub.co}/api/cli/commits" -d @payload.json
```
`payload.json` shape (build it with git + jq; keep `diff` under 1.5 MB):
```json
{
"projectId": "<uuid>",
"sha": "<full sha>",
"message": "<commit subject>",
"committedAt": "<ISO timestamp>",
"diff": "<git show output>",
"filesChanged": 3, "insertions": 42, "deletions": 7,
"conversation": {
"source": "claude-code",
"model": "<model id if known>",
"messages": [
{ "role": "user", "content": "..." },
{ "role": "assistant", "content": "..." },
{ "role": "tool", "toolName": "Edit", "content": "..." }
]
}
}
```
Roles must be `user | assistant | tool | system`; each message's `content` is
capped at 200k chars; max 500 messages per call. To reconstruct the full
transcript, read this session's JSONL transcript file and map assistant
`tool_use` blocks to `role: "tool"` messages with `toolName` set.
## Store agent output without the hooks
For ad-hoc output (a report, a build log, subagent results) not tied to a
commit, create a session and append messages:
```bash
# 1. create (idempotent on externalId)
curl -sf -X POST -H "Authorization: Bearer $VIBEHUB_TOKEN" -H "Content-Type: application/json" \
"${VIBEHUB_API_URL:-https://www.vibehub.co}/api/sdk/sessions" \
-d '{"source": "claude-code", "summary": "<what this output is>", "kind": "live"}'
# 2. append (repeat as needed)
curl -sf -X POST -H "Authorization: Bearer $VIBEHUB_TOKEN" -H "Content-Type: application/json" \
"${VIBEHUB_API_URL:-https://www.vibehub.co}/api/sdk/sessions/<SESSION_ID>/messages" \
-d '{"messages": [{"role": "assistant", "content": "<the output>"}]}'
```
## Projects
- List: `GET /api/cli/projects`
- Create (idempotent by slug): `POST /api/cli/projects` with
`{"name": "<repo name>"}` — use the repo directory name unless the user says
otherwise, and reuse the returned `id` as `projectId` in later calls.
## Reporting back
After a capture, give the user the session/commit id returned by the API and
the dashboard URL (`${VIBEHUB_API_URL}/dashboard`). On any 4xx, show the
response body — it contains a precise validation message. Never print the
token itself.
Versions
| Repository | Score | Licence |
|---|---|---|
| This page · aw3-technology/vibehub-plugins | 100 | MIT |
| wenlan-app/wenlan | 100 | Apache-2.0 |
| wenlan-app/wenlan | 100 | Apache-2.0 |
| optimiziramsi/skills | 100 | MIT |
| Kolibee-GmbH/m2b | 100 | MIT |
| davepoon/buildwithclaude | 100 | MIT |
| anurag-rivora/product-brain-starter | 100 | MIT |
| borghei/Claude-Skills | 100 | MIT |
| alirezarezvani/claude-skills | 100 | MIT |
Database snapshot: 13.11.2026 · package version da2a3a018dbf (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 92git-commit
Use when creating Git commits. - 96commit
Creates atomic conventional git commits using git-agent. This skill should be used when the user requests "commit", "git commit", "create commit", or wants to commit changes using AI atomic commits. - 100commit
Create well-structured git commits from the current working tree. Use when the user says 'commit', 'save my work', 'let's commit this', 'make a commit', or any variation of wanting to commit code to git. - 100git-commit
Create commits with conventional messages and issue references. Use when user says "commit", "save changes", or "stage and commit". Local commits only — see git-push for remote. - 100git-commit
Creates git commits following Conventional Commits format with type/scope/subject. Use when user wants to commit changes, create commit, save work, or stage and commit. Enforces project-specific conventions from CLAUDE.md. - 95commit
Creates well-structured git commits using Conventional Commits format. Activate this skill whenever the user asks to commit changes, make a commit, save progress, or says "commit this", "commit my changes", "/commit", or any variation of requesting a git commit. Also activate…