Project Retrospective

Team·Spring·3% of the spring grade

• Project RetrospectiveSunday of week 9·Rubric

Up to 48 hours late loses one rubric band on each criterion. After that it scores zero, unless you contacted the instructors before the deadline with a documented reason.

In this retrospective your team writes down the shape of three terms of work: the project’s arc, a few pivots, the decisions that aged well and the ones that turned into interest payments, while everyone who lived it is still in the room. It’s also the shared raw material each of you will mine for the individual Career and Individual Retrospective: write it well and the individual assignment gets easier for everyone.

Section 5 is addressed to whoever continues the work, so it also goes to your partner with the handoff document. The Retrospectives guide has prompts for the cross-year format.

What You Produce

Write 3 to 5 pages, each section under its own heading and every claim naming its example, kept with your team’s documents, not in the repository.

AI use: AI may help synthesize and draft from your own repo history, notes, and sprint records; that is a legitimate use of the paper trail you built. It must not invent events, soften consequences, or manufacture artifacts. Graders have read your sprint notes all year and will notice a story that does not match them.

1. Arc and Pivots

The story of three terms as a narrative, not a term-by-term status list: what you set out to build in fall week 1, the three or four milestones that actually mattered, where momentum was strong and where it stalled, and each significant pivot or descope with what triggered it.

2. Delivered Versus the Definition of Shipped

Put the final state next to your Definition of Shipped: the target rung, the metrics, and what was actually achieved (this should agree with your Release and Metrics numbers). Name the gap plainly and explain it.

3. Engineering Decisions and Their Consequences

Pick 3 to 5 significant technical decisions from the year and trace each to its consequence: what it enabled, what debt it created, whether you would make it again. Link the RFCs and ADRs in docs/rfcs/ or docs/adr/ where the decisions were recorded (the ADR guide covers the status field you will be reading a year later, and adr.github.io collects the templates); this is where the year’s paper trail pays off. At least one decision should be one that did not work out.

4. Team Dynamics

How the team actually functioned across the year: how the rotating roles evolved, what communication patterns helped or hurt, how conflict was handled when it came (referencing the charter’s conflict path if you used it), and what this team knows about working together that it did not know in fall week 1.

5. For Whoever Continues the Work

A forward-looking section addressed to the next owner, whether that’s your partner, a future capstone team, or open-source maintainers: what they should understand about the architecture before touching anything, which risks in docs/risks.md are still live (the Project Risks guide covers what a live entry carries), and the first three things you would do next. This section feeds directly into the Project Handoff.

Include at least two lightweight artifacts across the document (a pivotal PR, a before-and-after diagram, a board snapshot), each with a caption saying why it matters.

What You Submit

The retrospective as one PDF, with the cover Submitting Your Work describes.

Rubric

Rubric: 100 points across 6 criteria

Criterion	Ratings
Arc and pivots20 pointsOutcomes: SO3	Exceeds Expectations (20)Each term's milestones are dated and linked to their artifacts (a tag, a PR, a checkpoint), and each pivot names what was decided, when, and the evidence that prompted it. Meets Expectations (16)The milestones are there but some are undated or unlinked, or a pivot is mentioned without the decision, the date, or the evidence behind it. Does Not Meet Expectations (4)No cross-year arc, or an account that contradicts the team's sprint notes on a milestone or pivot.
Delivered versus shipped20 pointsOutcomes: SO2	Exceeds Expectations (20)A comparison against the Definition of Shipped that names the target rung, each metric, and what was achieved, with numbers that reconcile with the release report and any gap named and explained. Meets Expectations (16)The comparison is present but leaves out the target rung, a metric, or what was achieved, does not name and explain the gap, or its numbers do not reconcile with other submitted evidence. Does Not Meet Expectations (4)No comparison against the Definition of Shipped.
Decisions and consequences25 pointsOutcomes: SO6	Exceeds Expectations (25)Three to five decisions traced from RFC through to outcome, each naming what it enabled and the debt it created, including at least one that did not pan out. Meets Expectations (20)Decisions are traced, but a decision does not say what it enabled or the debt it created, a consequence is stated with no artifact behind it (an RFC, an ADR, a pull request, a metric), or every decision is presented as a success. Does Not Meet Expectations (5)Fewer than three decisions traced, or no link drawn between decision and outcome.
Team dynamics15 pointsOutcomes: SO5	Exceeds Expectations (15)Covers how roles evolved, which communication patterns helped and which hurt, and how conflict was handled, each tied to an example from the year. Meets Expectations (12)One or two of the three are not covered or have no example, or only what helped is described. Does Not Meet Expectations (3)Absent, or limited to "we worked well together".
Forward-looking section: understand first, live risks, first actions10 pointsOutcomes: SO2	Exceeds Expectations (10)Names what a successor must understand first; links the live risks in docs/risks.md; lists the first three actions to take. Meets Expectations (8)One of the three missing. Does Not Meet Expectations (2)Two or more of the three missing.
Writing and evidence10 pointsOutcomes: SO3	Exceeds Expectations (10)3 to 5 pages, each required section under its own heading, every claim names its example, and at least two captioned artifacts. Meets Expectations (8)One of the four missing. Does Not Meet Expectations (2)Two or more of the four missing.

Activities That Prepare This

Your spring Team Health Assessment feeds the team-dynamics criterion, and Learning Journal entries from all three terms are what make the arc-and-pivots criterion possible to write accurately.

Browse Learning and Reflection and Team and Workflow when these run out.