ditrixnew/edt-mcp/edt-mcp-project-runtime-debug

Claude Code and agent skill by ditrixnew/edt-mcp

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Collect bounded runtime and debugger evidence for a 1C project through EDT-MCP, including launch, breakpoints, variables, event log, and cleanup.

Install

npx vskill@latest install ditrixnew/edt-mcp/edt-mcp-project-runtime-debug

SKILL.md

EDT-MCP runtime debugging

Purpose and trigger

Use this skill when one runtime question needs authorized launch, Attach, debugger, variable, expression, or event-log evidence.

Operating rule

Read and apply the common operating rules before this workflow.

Task boundary

Resolve one exact project, application, launch/Attach configuration, and runtime question. Skip launch and debugger work entirely when bounded get_event_log evidence alone answers the question.

Primary workflow

  1. For event-log-only work, resolve the authorized log source, apply current

server-side bounds from help, call get_event_log, and report any paging, format, location, disclosure, or completeness gap.

  1. Otherwise resolve the target with get_applications,

list_configurations, debug_status, and list_breakpoints; distinguish pre-existing breakpoints from task-owned ones, and settle update, external-change, restart, credential, and data-disclosure authority before launching.

  1. Use set_infobase_credentials only for the confirmed target and only when

authorized. Decide from debug_status whether to use an existing session or start a fresh one before installing the smallest set_breakpoint. Use an existing session directly only when it is the exact authorized target and no fresh-start effect is required; install the breakpoint and continue without launch, or stop. For a fresh launch or Attach, set the breakpoint before starting it and retain task-owned identifiers. When that exact session is already running and no breakpoint line is known, debug_pause stops it at the next BSL statement it runs instead.

  1. If a preflight race makes launch report alreadyRunning=true, do not

claim that launch, update, restart, or startup options ran. Refresh the uniquely identified target, resume only a suspension caused by the task's breakpoint, remove task-owned temporary state, and stop unless a fresh relaunch was explicitly authorized. Never resume an ambiguous or unrelated suspension.

  1. Settle any in-progress or unknown launch outcome under current help and a

bounded caller-approved deadline before removing a task-owned breakpoint, treating the launch as absent, or completing. Before debug_pause, wait_for_break, get_variables, evaluate_expression, set_variable, step, or resume, require current help and status to identify one unambiguous intended debug target. Otherwise remove only task-owned temporary state and stop.

  1. Collect only the bounded evidence needed. Treat expression evaluation and

variable mutation as potentially state-changing.

  1. Resume execution if this task suspended it. A debug_pause that returned

paused: false stays armed: catch it with wait_for_break and resume it, or report the armed request. Call remove_breakpoint, and use terminate_launch only for a uniquely identified, task-owned launch whose termination is authorized.

Authority rule

Infobase update/restructure, external-change handling, credentials, restart, Attach, sensitive-data disclosure, expression/state mutation, and launch termination each require authority for the exact target and effect.

Stop rule

Stop on ambiguous debugger target, missing launch/Attach route, unapproved side effects or disclosure, shared-session risk, or uncontrolled BSL execution.

Completion signal

Report the exact runtime target and mode, decisive frame/value/log evidence, server-side bounds and partial-result caveats, any state mutation, and cleanup of task-owned suspensions, breakpoints, and launches.

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

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
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/ditrixnew/edt-mcp/edt-mcp-project-runtime-debug/badge)](https://verified-skill.com/skills/ditrixnew/edt-mcp/edt-mcp-project-runtime-debug)

Something wrong with this skill or its page? Tell us.