ops-sprint

Objective

Compute current sprint health: velocity, burndown, and completion prediction.

Before proceeding, verify that .cocoplus/ exists. If not, output: "CocoPlus is not initialized. Run $pod init first." Then stop.

Step 1 — Read Sprint Window Config

Read cocoplus.toml and extract [sprint] section:

• start_date — ISO date when current sprint began
• end_date — ISO date when current sprint ends (optional; compute 2-week default if absent)
• feature_prefix — commit tag prefix identifying feature work (default: feat)
• scope — optional list of pipeline/module names to scope analysis

If cocoplus.toml is absent or [sprint] section is missing, use defaults:

• start_date: 14 days ago
• end_date: today
• feature_prefix: feat

Step 2 — Compute Sprint Metrics Deterministically

Compute sprint-specific metrics from git log directly. Extract:

Story Velocity:

git log --oneline --since="<start_date>" --until="<end_date>" --grep="^feat"

Count commits matching the feature prefix. Each qualifying commit = 1 velocity unit.

Pipeline PRs (breadth signal):

git log --oneline --since="<start_date>" -- "*.sql" "*.py"

Count pipeline-touching commits as proxy for pipeline tasks completed.

Daily Burn Rate:

velocity_units / elapsed_sprint_days

Completion Prediction:

remaining_days = (end_date - today) in days
projected_additional = daily_burn_rate × remaining_days
predicted_final = velocity_units + projected_additional

Write results to .cocoplus/ops/sprint-health.json (atomic write via tmp→rename):

{
  "computed_at": "<ISO timestamp>",
  "sprint_window": {
    "start": "<start_date>",
    "end": "<end_date>",
    "elapsed_days": <N>,
    "remaining_days": <N>
  },
  "metrics": {
    "velocity_units": <N>,
    "pipeline_commits": <N>,
    "daily_burn_rate": <float>,
    "predicted_final_velocity": <float>
  },
  "signals": {
    "commits_in_window": <N>,
    "feature_prefix_used": "<prefix>"
  }
}

Step 3 — Spawn Haiku for Sprint Narrative

Pass sprint-health.json to a Haiku sub-agent with this mandate:

"Read the sprint health JSON and produce a concise narrative. Every statement MUST cite specific quantities or dates from the data. The narrative must:

1. State current velocity and daily burn rate
2. Project sprint completion based on trend
3. Identify whether pace is ahead, on track, or at risk
4. Suggest 1-2 prioritized actions if at risk

Do not produce vague generalities — cite numbers."

Step 4 — Commit sprint-health.json

Commit .cocoplus/ops/sprint-health.json to git only if the file has changed since the last commit (use git diff --quiet .cocoplus/ops/sprint-health.json to check). If unchanged, skip the commit and note: "Snapshot unchanged since last run — no commit needed."

git add .cocoplus/ops/sprint-health.json
git commit -m "ops(sprint): update sprint health snapshot <YYYY-MM-DD>"

Step 5 — Display Sprint Health Report

CocoOps Sprint Health — <project> — <date>

Sprint Window:   <start_date> → <end_date>  (<elapsed> / <total> days)

Velocity:        <velocity_units> units  (<daily_burn_rate>/day)
Pipeline PRs:    <pipeline_commits> commits
Completion:      Projected <predicted_final_velocity> units by <end_date>

Status: <ON TRACK | AHEAD | AT RISK>

<Haiku narrative — specific numbers, no vague generalities>

Snapshot: .cocoplus/ops/sprint-health.json

Anti-Rationalization Table

Shortcut / Temptation	Why It Fails
Let LLM compute velocity	Metrics must be deterministic — git log counts are evidence, not inference
Omit sprint window from config	Without a defined window, velocity numbers are meaningless
Skip committing sprint-health.json	Sprint snapshot is a team artifact — teammates must see the same numbers
Invent projected velocity without daily burn rate	Projection must use actual measured burn rate, not a guess

Exit Criteria

• Sprint window read from cocoplus.toml or defaults applied
• velocity_units, daily_burn_rate, predicted_final_velocity computed from git log
• sprint-health.json written and committed
• Haiku narrative cites specific numbers
• Report displayed in standard CocoOps format