Works with the AI Agents you already use
learning program effectiveness analyst
This skill converts raw, pseudonymous training records into a professional effectiveness review using a deterministic execution script for exact arithmetic.
$2.99
learning program effectiveness analyst
Example session with this skill installed
Analyze the effectiveness of our 'Q3 Leadership Sync' program using the attached CSV of pre/post scores and the revenue outcomes for participants and the control group. Max score is 100.
- Read your context and instructions
- Compiled the learning program effectiveness
- Generated the data export
I have processed the Q3 Leadership Sync records from the provided dataset (E1). The analysis was performed using the deterministic effectiveness script to ensure statistical accuracy.
Learning Program Effectiveness Review — Global Tech Corp, 2023-10-15
1. Scope, sources and assumptions
- Sources: E1 (Training_Export_Q3.csv).
- Programs: Q3 Leadership Sync (ID: LD-2023-Q3).
- Accounting: 42 enrollments processed; 5 records excluded due to missing consent.
- Assumptions: Revenue outcomes are attributed to the primary cost center of the learner (Source: E1).
2. Participation and completion
| Program | Enrolled | Started | Completed | Withdrew | In progress | No-show | Participation | Completion | Attrition |
|---|---|---|---|---|---|---|---|---|---|
| Leadership Sync | 37 | 35 | 32 | 2 | 1 | 2 | 0.9459 | 0.8649 | 0.0541 |
3. Skills improvement
| Program | Paired n | Mean pre | Mean post | Mean gain | Normalized gain | Effect size d_z | Share improved |
|---|---|---|---|---|---|---|---|
| Leadership Sync | 30 | 62.50 | 84.20 | 21.70 | 0.5786 | 0.82 | 0.9000 |
Normalized gain: The proportion of potential gain achieved. Effect size d_z: Standardized difference between paired observations.
4. Business outcomes and attribution limits
| Metric | Direction | Before | After | Change | Comparison Change | Diff-in-Diff |
|---|---|---|---|---|---|---|
| Revenue (USD) | Increase | 45,000 | 52,000 | +7,000 | +2,500 | +4,500 |
Attribution Limitations: Observed changes may be influenced by Q3 seasonal market shifts. This design uses a non-randomized comparison group; a causal claim is not supported.
5. Suppressed groups, exclusions and data gaps
- Suppressed: Qualitative feedback for 'Region: North' suppressed (n<5).
- Exclusions: 5 records removed for
consent=false.
Information gap: Data source for comparison group revenue (E2) lacks timestamp verification —
Information gap — organizational validation required.
6. Warnings and items for human review
W_OUTLIER_OUTCOME: Record R-992 shows a 400% revenue increase; human review recommended to ensure this wasn't a manual entry error.
7. Interpretation and draft actions
- Completion rate (86.49%) exceeds the internal 80% benchmark (Computed result).
- The effect size (0.82) indicates a large practical significance in skill acquisition (Computed result).
Analytical estimate — validation required: If the Diff-in-Diff holds, the program correlates with a 10% revenue lift over baseline.
- Proposed action: Expand the program to Q4 cohorts based on the high share of improved learners (Draft action).
Next steps
- Validate the revenue outlier in record R-992.
- Confirm the data source for the comparison group to bridge the information gap.
- Present the draft action to the HR committee for the Q4 rollout decision.
This review supports, and does not replace, professional judgment and management's decision.
Training_Export_Q3.csv
CSV · data export
Example file from a real run - the skill writes it into your workspace.
Connects securely to your tools. The creator never sees your data.
About this skill
Training evaluation often relies on gut feel or messy spreadsheets that ignore statistical significance and privacy. This skill converts raw, pseudonymous training records into a professional effectiveness review using a deterministic execution script for exact arithmetic.
What it does
- Participation tracking: Generates tables for enrollment, completion, withdrawals, and attrition rates across multiple programs.
- Skills gain analysis: Calculates paired pre and post score means, normalized gain, and Cohen's d_z effect sizes.
- Outcome mapping: Relates business metric changes to program timing and compares results against optional non-participant groups.
- Data privacy enforcement: Automatically suppresses results for small groups to protect learner anonymity and filters records lacking consent.
- Attribution scoping: Explicitly states limits of causality and identifies concurrent factors that may influence results.
How it works
- Data Preparation: You provide pseudonymous enrollment records, scores, and business outcomes.
- Schema Validation: The agent builds a structured
request.jsonensuring all required parameters like score maximums and access scopes are met. - Deterministic Execution: A Python script processes the arithmetic to ensure decimal accuracy and consistency.
- Result Interpretation: The agent maps the script output to a structured report, flagging warnings and suppressed cells.
Frameworks & tools
Python 3 for deterministic calculations, JSON Schema for input validation, and standard Markdown for reporting.
Why this beats prompting it yourself
Standard LLM prompts often hallucinate averages, round inconsistently, or ignore the statistical necessity of paired samples. This skill uses a hardened script for the math while the agent handles the complex task of data cleaning, privacy suppression, and attribution logic.
Use cases
- Evaluate the ROI of a mandatory compliance or technical upskilling program.
- Compare the effectiveness of different training vendors using paired skills testing.
- Analyze the link between a sales training program and actual quarterly revenue changes.
- Generate auditable program reports that respect organizational privacy thresholds.
Known limitations
Does not perform individual performance reviews or workforce gap planning. Requires pseudonymous data; cannot process records containing real names or personal identifiers.
Requires pseudonymous data; no real names. Minimum group size defaults to 5 for suppression. Does not support causal claims or individual performance ratings.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 2 days ago