In short:
ak-fable-thinkingmakes the AI name the source of every claim, check output constraints by machine or unit by unit, and lead its answer with the result. It picks its checks by how risky the task is, then answers; once it finds the cause of a bug it hands off toak-fix, and when options need more discussion it hands off toak-brainstorm. It is not a workflow with hand-off steps, and it does not promise better reasoning on every task.
Disclosure: I maintain AgentKit and have an affiliate relationship with it: the AgentKit link in this post is an affiliate link; buying through it gets you 30% off.
Summary
Use it when you need an answer that is right more than one that is fast: diagnosing a bug, a review, picking an architecture, a hard decision, or a passage that must hit an exact word count. You do not need it to look up a fact or for a small, easy-to-undo task: there it answers directly, and running deep checks only burns tokens.
flowchart LR
A["Task that must be right"] --> B["ak-fable-thinking"]
B --> C["Answer: result first, evidence, weak spots"]
B --> D["Text checked against its constraints"]
B --> E["Stuck several times: change approach"]Four concepts to know:
| Concept | Question | Example |
|---|---|---|
| Constraint Loop | Does the output have constraints a machine can count or scan? | "Exactly 50 words, no letter e" means the AI must count and scan every word, not reread it and decide it passes |
| Claim classification | Did the AI see this itself, derive it, remember it from training, or is it assuming? | "Library X supports Y" is only old memory until the AI opens the docs or runs it in this session |
| Check the real outcome | If you follow the answer exactly, is the goal actually reached? | The mail server is down: report it to IT by email or by phone? Email cannot reach IT, so the answer is to call |
| Depth | How deep is this task worth checking? | A lookup gets a direct answer. A production incident, a data migration or security gets careful checks plus the options that were ruled out |
Only a tool can raise a claim to "seen firsthand"; saying it more confidently cannot. "I don't know" plus a way to find out is also a valid answer.
Sample prompt for a flaky test on CI:
/ak-fable-thinking Test <test name> fails about 1 in 5 runs on CI.
Find the root cause, do not add retries or sleeps.
Done when: there is evidence of the cause and the test still means something.This skill has no flags.
Below is the detailed version: the run flow, 5 situations with prompts, the change history by ak release, and common pitfalls. The detailed version and the interactive page are for subscribers. Your agent can also search them through the portal's MCP.
Opens when your account has subscriber access
This guide is for subscribers
Once your account has access, the full piece, the interactive visual and the agent call all open.
- Sign in or create a modernEngineer accountReading access belongs to your account email.
- Choose a plan on the pricing page1, 3 or 12 months, paid by VietQR bank transfer. Access opens about a minute after the money arrives.
- Or redeem a Facebook subscriber codePeople supporting the page on Facebook get an activation code by message each month.
Already a subscriber? Sign in to read.
Use this piece in your agent
Get an API keyCopy the text below and paste it into Claude Code, Codex or whichever agent you use. The agent reads the piece and checks it against the repo you have open.
Read "ak-fable-thinking: make the AI check evidence before it answers" on modernEngineer and apply it to this repo.
1. Call the content_get tool of the modernengineer MCP server with locale "en", slug "ak-fable-thinking". If the modernengineer MCP server is not connected, stop and walk me through connecting it from https://modernengineer.dev/en/mcp.
2. Summarize the main steps of the piece and point out which apply to this repo.
3. Propose concrete changes and ask me before editing any file.Connect MCP in Claude Code (once)
claude mcp add --transport http modernengineer \
https://api.modernengineer.dev/mcp \
--header "Authorization: Bearer <API key>"Reading the full piece over MCP needs an API key from a subscriber account or one that bought this guide. Without access, the agent only gets the opening.
- Version
- v1.0.0
- Updated
- 22/10/2026
- Licence
- All rights reserved
- Slug
- ak-fable-thinking
