AI at Work

Send the made-up Community Skills Afternoon pack. Include an organized file list, event brief, budget with visible formulas and an unsent review email. Add your AI-assisted planning document, its sources and a short note about how you used AI. Show at least one AI mistake you checked and corrected. Explain one calculation. Include the readable work itself; a link or claimed accuracy score is not enough. Attach readable code and text files. A link or claimed test result alone is not enough. If important work is missing or unclear, we may ask for more or send it to a school reviewer.

• Source and artifact brief (35%) · essential
  The readable static artifacts cover the authored workplace and AI-planning tasks.
• Source and design decisions (30%) · essential
  Visible calculations, cited sources, attribution and corrections support the stated decisions using fictional data.
• Source analysis and check design (25%) · essential
  Show a source-backed correction, a representative manual calculation and limitations; distinguish documented reasoning from an independently verified software result.
• Explanation, attribution and revision (10%)
  Explain a specific source/artifact decision and revision, attribute assistance and state the limits of what the submitted evidence establishes. Unresolved evidence gaps require clarification or optional human review.

Make a small Python program that adds up study time. Send it as solution.py. Read exactly one JSON value from stdin: an array/list of at most 200 fictional {topic:string, minutes:integer 1..600, completed:boolean} records. For valid input, write exactly one JSON object to stdout: {totalMinutes:sum of all minutes, completedCount:number of completed sessions, minutesByTopic:object of minute totals keyed by each exact topic string}, then exit with status 0. An empty array/list returns zeros and an empty object. For invalid input, exit with status 2 and no stdout: reject non-arrays/lists, more than 200 entries, malformed entries, non-string or blank/whitespace-only topics, non-integer/out-of-range minutes and non-boolean completed values without coercion. Validate topic.strip() but preserve each original exact topic string as the grouping key. Python booleans are not valid integer minute values. Use no external dependencies, network or file access except reading stdin. Do not print debug messages or a test/pass report. You may organize the script with helper functions; the school assesses actual stdin/stdout behavior in fresh processes, comparing outputs independently on the host. Include a short explanation with your submission. Passing this bounded suite does not establish correctness of the entire capstone.

• Correct summary (35%) · essential
  Write the exact JSON summary for valid stdin input and exit 0; totals include every session, completed counts and original-topic totals.
• Input validation (25%) · essential
  Reject invalid lists, entries and fields with exit status 2 and no stdout, without silently coercing strings, numbers or booleans.
• Trusted test results (25%) · essential
  The isolated trusted suite must actually execute and pass; unavailable execution is not a pass.
• Explanation (15%)
  Explain the function, representative checks and one error corrected.

Send the event-report generator code. Include made-up input, an example report, the export format and your plan for a dry run and bad input. Add a text copy of a Git change and your response to review. Follow normal input, bad input and a dry run through the code. Explain how the code protects the original files. A supplied log does not prove that the code or its checks ran. Attach readable code and text files. A link or claimed test result alone is not enough. If important work is missing or unclear, we may ask for more or send it to a school reviewer.

• Source and artifact brief (35%) · essential
  The source and change artifacts cover the authored report generation, export and Git-review brief.
• Source and design decisions (30%) · essential
  Input validation, dry-run behavior and file-boundary decisions are explicit and internally coherent in the code.
• Source analysis and check design (25%) · essential
  Trace ordinary, invalid and dry-run paths to source locations, design checks and explain a review-driven correction without relying on a claimed pass report.
• Explanation, attribution and revision (10%)
  Explain a specific source/artifact decision and revision, attribute assistance and state the limits of what the submitted evidence establishes. Unresolved evidence gaps require clarification or optional human review.

Send the project case study, a practice resume for a stated job and a short written presentation. Use made-up personal details. Support your project claims with the code or documents you send. Explain your own work, any AI help and one correction you made. We review these materials; we do not judge your job prospects or verify that the software runs. Attach readable code and text files. A link or claimed test result alone is not enough. If important work is missing or unclear, we may ask for more or send it to a school reviewer.

• Source and artifact brief (35%) · essential
  The case study, resume draft and presentation address the actual portfolio brief.
• Source and design decisions (30%) · essential
  Claims accurately reflect supplied artifacts and disclose AI assistance and limits.
• Source analysis and check design (25%) · essential
  Trace project decisions and claimed learning to readable evidence and a documented correction; avoid unverified performance claims.
• Explanation, attribution and revision (10%)
  Explain a specific source/artifact decision and revision, attribute assistance and state the limits of what the submitted evidence establishes. Unresolved evidence gaps require clarification or optional human review.

Combine your checked event documents, Python reporting code, Git changes and project presentation into one handover. Use made-up details. Follow a changed input, bad input and a dry run through the code. Show a correction and explain how another person could repeat the work. The separate small Python task has school code checks. This final project is reviewed by reading the code and documents. Attach readable code and text files. A link or claimed test result alone is not enough. If important work is missing or unclear, we may ask for more or send it to a school reviewer.

• Source and artifact brief (35%) · essential
  The integrated handover uses actual digital, AI, Git, Python-automation and portfolio artifacts from all six courses.
• Source and design decisions (30%) · essential
  The supplied document/source decisions handle attribution, checked information, validation and file protection coherently.
• Source analysis and check design (25%) · essential
  Source-linked input-change/error/dry-run traces, expected results and revision evidence support the explanation without claiming the complete workflow was executed.
• Explanation, attribution and revision (10%)
  Explain a specific source/artifact decision and revision, attribute assistance and state the limits of what the submitted evidence establishes. Unresolved evidence gaps require clarification or optional human review.