VespeR — Development agent for Claude Code
Control plane for Claude Code workflows: interactive sessions, agents, patterns, DAGs, and resumable context.
How to install VespeR
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open ranausmanai/VespeR and copy the folder into
~/.claude/agents/.
What VespeR does
Control plane for Claude Code workflows: interactive sessions, agents, patterns, DAGs, and resumable context.
Alternatives in Development
- Access Control — PostHog access control system implementation expert - use when adding access controls to new products, debuggi 39.2k ★
- On Call Handoff Patterns — Effective patterns for on-call shift transitions, ensuring continuity, context transfer, and reliabl 31.9k ★
- Claude Code Tools — by Prasad Chalasani - Well-crafted toolset for session continuity, featuring skills/commands to avoid compacti 1.6k ★
README
VespeR
Visual control plane for Claude Code workflows.
[Website](https://vesper.run) • [Repository](https://github.com/ranausmanai/VespeR)
VespeR helps you run coding workflows with more control than a single chat stream:
- reusable specialist agents
- reusable multi-agent patterns
- execution observability (DAG + timeline + activity)
- resumable interactive sessions with smart memory context
CLI commands:
vesper(preferred)agentling(backward-compatible alias)
Product Tour


What VespeR Does
- Interactive sessions you can rejoin
- Start multiple interactive sessions and switch between them.
- Stop current response without ending the session.
- Agent Dock in chat
- Ask a specialist agent (reviewer, security, strategist, etc.).
- Inject agent output back into the active interactive session.
- Pattern execution
- Run saved patterns such as:
Quick Build (Solo)Build + Review LoopRelease Readiness PanelExpert PanelDebate
- Execution visibility
- See what happened via DAG, timeline, and split activity views.
- Inspect tool usage and file-touch activity during agent/pattern runs.
- Smart resume memory
- Ended runs are converted into structured memory.
- New interactive runs can start with ranked context packs (goal, touched files, open loops, validations, next action).
Key Use Cases
- Build and review in one pass
- Run
Build + Review Loopso code generation and code review are linked.
- Release go/no-go checks
- Run
Release Readiness Panelfor security/performance/product sign-off.
- Long-running continuity
- End and later resume with context pack instead of replaying long transcripts.
- Debugging agent behavior
- Use DAG/timeline/activity to diagnose loops, failures, and churn.
Requirements
- Python
3.10+ - Node.js
18+ - Claude Code CLI available on
PATH
Install
# Clone
git clone https://github.com/ranausmanai/VespeR.git
cd VespeR
# Backend
pip install -e .
# Frontend
npm -C frontend install
npm -C frontend run build
# Native macOS shell (optional)
pip install -e ".[desktop]"
# macOS app bundle + DMG build toolchain (optional)
pip install -e ".[desktop-build]"
Run
# Start native macOS app
vesper desktop
# Start UI + API server in browser
vesper ui --port 8420 --no-browser
Open the browser UI at: `http://127.0.0.1:8420`
If `vesper` is not found, use:
agentling desktop
# Start UI + API server
agentling ui --port 8420 --no-browser
For frontend development inside the native shell:
vesper desktop --dev --frontend-url http://127.0.0.1:5173
Run `npm -C frontend run dev` in a second terminal first.
macOS App Distribution
Build a standalone `.app` and `.dmg` while keeping the browser UI flow intact:
pip install -e ".[desktop-build]"
bash scripts/build_macos_app.sh
Artifacts:
Related Agents
Promptplus Precommit Check
Use to catch, on pending code changes only, the two ConsolePlus/PromptPlus bug-risk patterns that are cheapest
Fpm Orchestrator
Master agent for FangPostMeet pipeline. Coordinates 6 sub-agents to process recordings end-to-end. Two phases
Lane Manager
Starts, resumes, parks and finishes lanes (branch + worktree + registry file) in the control plane. Use when t
De Airflow Expert
Expert Apache Airflow developer for DAG authoring, testing, and debugging. Use for DAG files (.py in dags/), a
Jdb Session.Agent
Run interactive JDB debugging sessions. Use when launching a JVM under JDB, attaching to a running JVM with JD
C2 Detection
Flag likely command-and-control / beaconing patterns.
Related Skills
SlackAgentBridge
Control local Claude Code and Codex CLI sessions from Slack with private channels and resumable terminals.
Harness Remote
Native-session control plane for Codex, Claude Code, OpenCode, OMP and PI. Run, resume and hand off coding ses
Observer
Local-first control plane for AI coding agents — launch browser terminals, run and remotely control sessions,