esonhugh/marketplace/terminal-session-debugging
Claude Code and agent skill by esonhugh/marketplace
- community
Checking that the SKILL.md file is still in the repository.
Use when debugging interactive CLI, REPL, TUI, ssh, telnet, nc, watch-mode, dev server, long-running command, terminal prompt flow, or any command that may not exit. Provides a PTY-backed MCP workflow for spawning terminal sessions, sending text and keys, reading output, managing multiple sessions, and exporting full transcripts.
Install
npx vskill@latest install esonhugh/marketplace/terminal-session-debuggingSKILL.md
Terminal Session Debugging
Use the terminal-session MCP server for terminal programs that need real PTY behavior or may not exit.
When to Use
Use this workflow for:
- Interactive CLIs, REPLs, shell prompts, installers, wizards, and prompt loops.
- TUI/curses programs where terminal size and key handling matter.
ssh,telnet,nc, and other bidirectional terminal/network sessions.- Watch-mode tools, dev servers, and commands that may run forever.
- Multi-session debugging such as server + client, listener + connector, REPL + shell.
- Any task where a normal Bash call may block, hide prompts, or lose interaction history.
Core Workflow
- Call
health_checkto confirm the MCP server is available. - Call
spawn_terminalwith a clearlabel. - Call
read_terminalto observe initial output. - Send one input at a time with
send_textorsend_key. - Read after every input before deciding the next action.
- Use
terminal_statusandlist_terminalsto inspect state. - Use
resize_terminalwhen terminal dimensions matter. - Use
close_terminalto interrupt, send EOF, terminate, or kill sessions. - Call
export_transcriptbefore reporting final results.
Recording Model
The MCP records command, output, stdin text, key events, resize events, close events, exit status, and errors under .terminal-debug/. It does not redact, mask, block, or safety-filter commands or content. Treat transcripts as complete debugging evidence.
Interaction Discipline
- Do not send a long script of answers until the prompt sequence is known.
- For prompt flows, treat each prompt as a state; capture, send one response, capture again.
- For multiple sessions, use labels such as
server,client,listener,repl, orssh. - Never claim a behavior happened unless it appears in
read_terminal,terminal_status, or the exported transcript.
Report Format
When reporting results, include:
- Session IDs and labels.
- Commands spawned.
- Inputs and keys sent.
- Observed output evidence.
- Interpretation of what the evidence proves.
- Transcript export paths.
- Cleanup status for active sessions.
Security
- Security scanPassed, 100 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingplugins/terminal-session-mcp/skills/terminal-session-debugging/SKILL.md
Popularity
- 7-day trend score
- 5
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/esonhugh/marketplace/terminal-session-debugging)