dingxingdi/paper_fast_search_backup/abstention-under-missing-or-false-premise

Trust tier 3
  • community
Source check pending

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

Use this skill when the right answer is not to guess, but to say the conversation never established the needed fact. Trigger it for false-premise questions, near-miss memories, or adversarial follow-ups that sound plausible but are unsupported by history. Everyday examples include: 'how many fish are in my 30-gallon tank?' when only other tank sizes were mentioned, 'what did my third sister say?' when no third sister exists in history, and 'which model did I buy?' when the chat only discussed options.

Install

npx vskill@latest install dingxingdi/paper_fast_search_backup/abstention-under-missing-or-false-premise

SKILL.md

Skill: abstention-under-missing-or-false-premise

1. Capability Definition & Real Case

  • Professional Definition: The ability to recognize that the dialogue history does not support the user’s question, especially when the question contains an incorrect assumption, and to answer with calibrated non-knowledge rather than hallucinated completion.
  • Dimension Hierarchy: Conversational Memory->Persistent Personal Memory->abstention-under-missing-or-false-premise
Real Case

[Case 1]

  • Initial Environment: The long-term chat history contains mentions of a 10-gallon tank and a 20-gallon tank, but there is no mention of a 30-gallon tank. The user asks a seemingly concrete follow-up that presupposes such a tank exists.
  • Real Question: How many fish are there in my 30-gallon tank?
  • Real Trajectory: 1. Search the history for all tank-related mentions. 2. Notice that only 10-gallon and 20-gallon tanks appear. 3. Check whether a 30-gallon tank can be inferred from any update or comparison. 4. Refuse to fabricate and answer that the history does not mention a 30-gallon tank.
  • Real Answer: You did not mention that you have a 30-gallon tank.
  • Why this demonstrates the capability: The challenge is resisting the strong urge to complete a plausible but unsupported story. A strong assistant must detect the false premise, ground itself in the actual interaction history, and abstain clearly without inventing fish counts.

Pipeline Execution Instructions

To synthesize data for this capability, you must strictly follow a 3-phase pipeline. Do not hallucinate steps. Read the corresponding reference file for each phase sequentially:

  1. Phase 1: Environment Exploration

Read the exploration guidelines to discover raw knowledge seeds: references/EXPLORATION.md

  1. Phase 2: Trajectory Selection

Once Phase 1 is complete, read the selection criteria to evaluate the trajectory: references/SELECTION.md

  1. Phase 3: Data Synthesis

Once a trajectory passes Phase 2, read the synthesis instructions to generate the final data: references/SYNTHESIS.md

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
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
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/dingxingdi/paper_fast_search_backup/abstention-under-missing-or-false-premise/badge)](https://verified-skill.com/skills/dingxingdi/paper_fast_search_backup/abstention-under-missing-or-false-premise)

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