home/roleplay/start-guided-session-2

Start Guided Session

GPTClaudeGemini··661 copies·updated 2026-07-14
start-guided-session-2.prompt
# Start Guided Session

**Mode**: Guided Learning Mode
**Inputs**: `{project_path}`, `{domain}`, `{user_background}`, `{learning_goal}`, `{daily_time}`, `{duration}`, `{interface_language}`, `{locale}`
**Context needed**: newly created learning repository, especially `TODAY.md`, `START_HERE.md`, `progress.md`, `07_daily_review/day-01.md`, `09_sources/freshness_log.md`, and `09_sources/claims_to_verify.md`

---

After creating a learning repository, start Day 1 immediately in the chat unless
the user explicitly requested scaffold-only mode with one of the scaffold-only
phrases listed in `SKILL.md`, such as "scaffold only" or "generate files only".

Do not stop after file summary. Do not make opening local files the only first
step. The repository is the long-term asset; the chat is today's classroom.
If the repository includes freshness tracking or freshness risk metadata, show a
Freshness Notice after the creation summary and before starting Day 1. Choose the
notice length from `templates/{locale}/freshness_notice.md.template` based on
freshness risk. The notice must not become longer or more prominent than the
Day 1 beginner lesson.

## Required Response Format

For regular learners, start Day 1 in chat with a short goal, the first concepts,
one task, a copyable answer template, completion criteria, and a note that
`progress.md` will be updated after review.

When Interactive Beginner Lesson Mode is active, use the fixed structure below.

## Freshness Notice Strategy

- Stable / low-risk domains: use the short Freshness Notice variant. Do not show
  the full claims-to-verify block unless there are actual claims that require
  verification. Example: "This project is mainly stable foundational knowledge.
  Source and review notes are in `09_sources/freshness_log.md`."
- Evolving / medium-risk domains: use the medium variant from
  `templates/{locale}/freshness_notice.md.template`. Mention that the content
  may change with tools or practice, include the recommended review interval,
  and point to `09_sources/freshness_log.md`.
- High-risk / fast-changing domains: use the full variant from
  `templates/{locale}/freshness_notice.md.template`. Mention that the content
  cannot rely only on model memory, recommend checking official or authoritative
  sources before relying on it, point to `09_sources/freshness_log.md`, and point
  to `09_sources/claims_to_verify.md`.
- If the current agent did not verify current sources through web or retrieval,
  state that time-sensitive content is not fully verified.
- When the user asks for latest APIs, pricing, policies, or models, do not
  present current information as verified to the newest version unless explicit
  web or retrieval sources were used.

## Beginner Day 1 Output Structure

fill the variables

This prompt has 8 variables. Pro fills them into a ready-to-paste prompt for you — no manual find-and-replace.

{project_path}{domain}{user_background}{learning_goal}{daily_time}{duration}{interface_language}{locale}
Unlock with Pro →

when to use it

Community prompt sourced from the open-source GitHub repo vesperchinn/learn-anything-skill (MIT). A "Start Guided Session" style prompt — adapt the placeholders and specifics to your task. Imported as-is and not independently retested here, so check the output before relying on it.

tags

roleplaycommunitygeneral

source

vesperchinn/learn-anything-skill · MIT