一键在 Manus 中运行任何 Skill

orchestrator-failed

星标0

分支0

更新时间2026年4月23日 17:46

Protocol for handling resume orchestrator failures. STRICTLY prohibits completion of failed jobs and mandates user-driven corrective actions.

安装

用 Codex 或 Claude 帮你安装复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。

在 Manus 中运行

来源

adamoutler

adamoutler/gemini-cli-resume-framework

打开 GitHub 仓库查看创作者相关仓库

下载

在 Manus 中运行

Orchestrator Failure Protocol

This skill is activated when the resume orchestrator script fails to complete successfully.

Context

A failure indicates that the system has attempted to generate and validate the resume 5 times in a row and has failed to reach a passing state. This suggests one of the following:

Conflicting Information: The generated content contradicts the source of truth.
Incomplete Information: The source data lacks the necessary evidence to support the claims.
Flawed System: The Personas (Builder, Auditor, Reviewer) or the Orchestrator script are malfunctioning or misaligned.

Core Mandate

You must NEVER complete the job if the orchestrator fails.

Procedure

Analyze Logs: Read the execution logs to determine the specific cause of failure (e.g., persistent factual inconsistencies, missing keywords, schema violations).
Trace Root Cause: Identify the source of the problem.
- Hallucination: Did the LLM invent facts?
- Input Data: Is the CV data missing required skills or experience?
- Persona: Is a persona instruction too strict, ambiguous, or incorrect?
- Script/Orchestration: Is there a bug in the automation logic?
Determine Least Impactful Solution: Evaluate potential fixes and their side effects (e.g., weakening validation rules vs. fixing data).
Present Options to User: You must not take proactive corrective action. Instead, present the findings and example options to the user.
- Option 1: Update CV Data (e.g., Add missing skills or evidence to the source files).
- Option 2: Update Persona/Reviewer Instructions (e.g., Relax constraints or clarify prompts).
- Option 3: Update Orchestration Logic (e.g., Fix bugs in the script or workflow).
Await Decision: Take no action until the user selects a path.

It is better to produce no resume and make no changes than to give the user a flawed resume, or potentially cause flaws in all future resumes.

同仓库更多 Skills

同仓库

cv-data-management

adamoutler/gemini-cli-resume-framework

CV Building Mode. Guidelines for adding new professional experience, projects, or skills to the cv-data repository.

2026-04-280

interview-intelligence

adamoutler/gemini-cli-resume-framework

Interview Preparation Mode - Generates a strategic "Intelligence Pre-Interview Sheet" by synthesizing the Job Description, Company news, and Candidate history. Use this when preparing for an interview or researching a target role.

2026-04-280

project-conventions

adamoutler/gemini-cli-resume-framework

General project rules and architectural documentation for AGENT_Tasks. Use this when modifying files in AGENT_Tasks, debugging the resume orchestrator, or editing automation scripts.

2026-04-280

resume-formatting

adamoutler/gemini-cli-resume-framework

Guidelines and constraints for JSON Resume generation and rendering. Use this when editing or validating resume data.

2026-04-280

resume-orchestration

adamoutler/gemini-cli-resume-framework

Execution mode - Workflow for generating tailored resumes and cover letters based on a Job Description (JD). Use this when the user asks to create a resume for a specific job or provides a position description.

2026-04-280

ats-optimization

adamoutler/gemini-cli-resume-framework

Strict rules for ensuring generated resumes are parsable by Applicant Tracking Systems (ATS).

2026-04-230

Orchestrator Failure Protocol

This skill is activated when the resume orchestrator script fails to complete successfully.

Context

A failure indicates that the system has attempted to generate and validate the resume 5 times in a row and has failed to reach a passing state. This suggests one of the following:

Conflicting Information: The generated content contradicts the source of truth.

Incomplete Information: The source data lacks the necessary evidence to support the claims.

Flawed System: The Personas (Builder, Auditor, Reviewer) or the Orchestrator script are malfunctioning or misaligned.

Core Mandate

You must NEVER complete the job if the orchestrator fails.

Procedure

Analyze Logs: Read the execution logs to determine the specific cause of failure (e.g., persistent factual inconsistencies, missing keywords, schema violations).

Trace Root Cause: Identify the source of the problem.

Hallucination: Did the LLM invent facts?
Input Data: Is the CV data missing required skills or experience?
Persona: Is a persona instruction too strict, ambiguous, or incorrect?
Script/Orchestration: Is there a bug in the automation logic?

Determine Least Impactful Solution: Evaluate potential fixes and their side effects (e.g., weakening validation rules vs. fixing data).

Present Options to User: You must not take proactive corrective action. Instead, present the findings and example options to the user.

Option 1: Update CV Data (e.g., Add missing skills or evidence to the source files).
Option 2: Update Persona/Reviewer Instructions (e.g., Relax constraints or clarify prompts).
Option 3: Update Orchestration Logic (e.g., Fix bugs in the script or workflow).

Await Decision: Take no action until the user selects a path.

It is better to produce no resume and make no changes than to give the user a flawed resume, or potentially cause flaws in all future resumes.

name	orchestrator-failed
description	Protocol for handling resume orchestrator failures. STRICTLY prohibits completion of failed jobs and mandates user-driven corrective actions.