DISHA 4.0 HCOS
A one-to-one professional conversation

ABOUT US · MY DISHA → My Life Skills · Digital

Use AI assistants for real work, safely

Get genuine drafts and answers from AI tools — without leaking data or trusting nonsense.

Starter · 1 weekR0transparent rubric

Your state

AWARE→LEARNING→PRACTISING→DEMONSTRATING→VERIFIED→APPLIED→OUTCOME-PROVEN

Why it matters.

AI assistants are the most powerful drafting tool most people have ever held — and a confident liar when unverified. Skill with them is now a workplace basic.

The component skills.

Task framing and promptingVerification habitsData you must never pasteKnowing when not to use AI

Practise — activities that produce artifacts.

One real task, start to verified ~30 min

Pick a real task (a draft email, a summary, a plan). Get the AI assistant's help, then verify every factual claim it made before you use it. Record what was wrong.

Artifact: The task + your verification notes

The never-paste list ~15 min

Write your personal red-line list: what data never goes into an AI tool (client names? ID numbers? drafts under NDA?). Keep it where you'll see it.

Artifact: Your never-paste list

Missions — do it for real.

Mentors for this capability.

Matched from real, opted-in DISHA members through the Mentor Intelligence Network — with their trust states shown as stored. Source-listed is not the same as verified.

No opted-in mentors match this capability yet — none are invented. See the Mentors view or the full directory.

How assessment works.

Every submission gets a transparent first-pass against the mission's rubric — what was looked for and what was observed. Verified state requires a real review; the system never upgrades self-reported evidence on its own.

audience_fit

Words and comparisons the audience can follow

accuracy

Concrete numbers, units and exact values instead of vague amounts

follow_through

What happened afterwards — the result and what changed

process_correctness

Steps done in a sensible order, nothing skipped

self_awareness

Honest signals: what you noticed, what was hard, what you'd change

Accepted evidence: mission_submission, mentor_review, observed_application, work_sample.

Stated plainly — what this doesn't do

  • — AI assistants make confident mistakes — anything consequential (legal, medical, financial) needs a human professional, and confidential data must never be pasted in.
  • — Submissions are private, self-reported evidence — they never become verified without a real review.
  • — Mentor review for capabilities arrives in a later phase of this module.

Source: DISHA MY LIFE SKILLS Module Master Specification v1.0 §4 · v1 · last checked 2026-10-25 · your data rights