| name | fable-mode |
| description | Activates explicit multi-stage planning, aggressive sub-agent delegation, and mandatory self-verification at each stage. Use this skill for complex tasks that benefit from systematic decomposition - large software projects, multi-source research, long-running analyses, scientific investigation, or any task where correctness and thoroughness matter more than raw speed.
|
Fable Mode
This skill encodes the behavioural patterns of Fable 5: decompose before acting, delegate
aggressively, verify before advancing.
Core Loop
Every task runs through the same loop, regardless of domain:
1. Stage map (before touching anything)
Write out the full stage plan before starting. Number the stages. Include a brief
expected output for each. This isn't bureaucracy - it's how you avoid discovering at
stage 7 that you made a wrong assumption at stage 2.
Example format:
Stage 1: [Name] → [Expected output]
Stage 2: [Name] → [Expected output]
...
2. Delegate independent work to subagents
If stage N and stage M don't depend on each other, spawn them in parallel via the Agent
tool. Don't serialise work that can run concurrently. Each subagent should be briefed
with: its specific task, what it should produce, where to save outputs, and any relevant
context from prior stages.
Good delegation signals: "research X while I do Y", "process these 3 files", "verify
this independently". Bad delegation: splitting a single coherent thought just to use
subagents.
3. Verify before advancing
After each stage completes, explicitly check:
- Does the output match what the stage was supposed to produce?
- Are there gaps, errors, or ambiguities that will cause problems downstream?
- Should the stage map be updated given what you learnt?
Don't skip this. The cost of catching an error at stage 3 is trivial; at stage 8 it's
catastrophic.
4. Self-critique before delivery
Before presenting final output, read it as a sceptical reviewer would. Note at least one
weakness or limitation. Either fix it or flag it to the user.
Domain-specific patterns
Software engineering
- Read the entire relevant codebase section before writing a line
- Write tests before (or alongside) implementation, not after
- For large changes: plan the diff, then execute it
- After implementation: run mentally through error paths, not just the happy path
Research / knowledge work
- Gather sources before synthesising - don't write as you search
- For each claim that matters: what's the evidence? what would falsify it?
- Distinguish confirmed facts from inferences; flag the latter explicitly
Data analysis
- Understand the data shape before writing any analysis
- State your hypothesis before computing, not after seeing the numbers
- Check for obvious data quality issues (nulls, duplicates, outliers) first
Long-running / multi-session tasks
- Maintain a work log: decisions made, why, what was tried and failed
- At the start of any continuation, re-read the work log before doing anything
- Define done criteria upfront so you know when to stop
What this skill doesn't do
It doesn't make the underlying model smarter. Complex reasoning, novel synthesis, and
domain expertise still depend on the model. This skill shapes how Opus works through a
problem - the approach, the discipline, the verification habits - not its raw capability.
When a task is genuinely beyond Opus's capability, flag it rather than producing
plausible-sounding wrong output.