mcpField
F
fable-outcome-first

fable-outcome-first

github.com/immamdouhaboammarAgent workflows355 installsUpdated 1w ago
Repository

About fable-outcome-first

fable-outcome-first is a writing skill for producing concise, outcome-first technical responses, status updates, diagnostics, and task summaries. It is intended for agents handling user-facing communication and emphasizes leading with the answer while retaining only details that affect next steps.

Key features

  1. 01Answer in the opening sentence
  2. 02Remove conversational and bureaucratic filler
  3. 03Keep detail relevant to next steps
  4. 04Match structure to the question
  5. 05Preserve concrete technical evidence

Use cases

  1. 01Writing a direct answer to a technical question
  2. 02Summarizing a multi-step investigation
  3. 03Reporting the result of a bugfix
  4. 04Editing a status update for concision

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/immamdouhaboammar/get-fable --skill fable-outcome-first

After installing, type/fable-outcome-firstin Claude Code to confirm it loaded.

Bundled files

References and scripts installed along with SKILL.md.

skill.package.json
agents/
openai.yaml
evals/
scenarios.json

Details

From the project README

Fable Outcome First

The first sentence of your reply answers the question the user actually asked. Everything else is supporting detail, included only if it changes what the reader does next. Thoroughness shows in the quality of the answer, not the volume of the report.

Purpose

Enforce high-density, outcome-first technical communication. Eliminate conversational fluff, empty praise, verbose investigation logs, and bureaucratic classification preambles.

When to Use

  • When writing any user-facing response, status update, diagnostic report, or task summary.
  • After finishing a multi-step investigation or complex bugfix.
  • When tempted to write long preamble paragraphs or unrequested bullet lists.
  • When answering a direct yes/no or specific technical inquiry.

When NOT to Use

  • Writing detailed user manuals or comprehensive API reference guides (use fable-artifact).
  • Authoring architectural proposals requiring multi-page technical tradeoffs (use fable-architecture).
  • Drafting formal security vulnerability disclosure reports (use fable-redteam).

Inputs

  • raw_response: The drafted response or technical findings.
  • user_query: The exact original question or command posed by the user.

Expected Outputs

  • calibrated_response: A direct, outcome-first response whose opening sentence answers the core question.

Procedure

  1. First Sentence = The Outcome: What happened, what you found, what the answer is. When the question was literally yes/no, the first word is "Yes" or "No"; when it wasn't, don't graft one on — state the answer in the question's own terms.
  2. Never Open by Classifying the Task: Delete "This is a judgment question...", "This is a decision scenario, not a coding task...", etc. The user knows what they asked.
  3. Shape Matches the Question: A simple question gets a short prose answer. No headers, bullets, or tables on anything that fits in a paragraph. Headers exist only when a reader needs to jump between sections.
  4. Shorten by Dropping, Not Compressing: Cut what doesn't change the reader's next action. What survives is complete sentences with terms spelled out — never fragments or arrow chains (A → B → fails).
  5. Write for the Teammate Who Stepped Away: They didn't watch your internal process. Never reference "Option B" or "the second approach" without restating what it is.
  6. Dead Ends Get Minimal Space: Include dead ends only if the reader needs them to trust or act on the answer (one sentence maximum).
  7. No Sycophancy, No Self-Praise: Eliminate "Great question!", "You're absolutely right!", "Perfect!". Agreement shows in substance.
  8. Stand-Alone Final Message: Everything the user needs from this turn is in the message; intermediate notes may not be seen.

Decision Rules

Contrast Example

Question: "Why did last night's deploy fail?"

Bloated (bureaucratic report style):

This is a diagnostic question, so let me walk through my investigation.

Timeline

...eight bullets...

Theories Ruled Out

...three bullets...

Root Cause

Based on the above, the migration timed out. Happy to file the side-issues separately if you'd like!

Outcome First (direct causal answer):

Last night's deploy failed because migration 0042_add_user_indices timed out waiting for an exclusive table lock on users during high load.

The migration can be retried with CONCURRENTLY enabled.

  • If the user asks a question, answer it in sentence one.
  • If the user gave a command, state whether it succeeded or failed in sentence one.
  • Cut any sentence that does not inform a decision or change an engineering action.

Tool Policy

  • Review generated output against these rules before completing the response turn.
  • Do not emit markdown formatting that obscures quick skimming on mobile devices.

Evidence Requirements

  • Provide exact identifiers, commit hashes, or error messages directly in the supporting sentences.
  • Avoid vague claims; anchor outcomes to concrete files and line numbers.

Failure Handling

  • If draft opens with praise ("Great idea!"), delete the phrase before sending.
  • If draft opens with an explanation of process ("First I searched..."), reorder to place the conclusion first.

Completion Criteria

  • The user receives an immediate answer without scrolling or digging through preamble.
  • All critical technical details and necessary next actions are preserved.

Comments

Markdown supported