DEVEX-01] Measure and improve developer experience with regular DevEx check-ins

[paruff]

[DEVEX-01] Measure and improve developer experience with regular DevEx check-ins

GitHub Issue Title: [DEVEX-01] Measure and improve developer experience with regular DevEx check-ins

Field	Value
Phase	Phase 4 — Quality Flywheel
DORA 2025 Source	“The platform capability most correlated with positive user experience is giving clear feedback on the outcome of tasks”; “Regularly survey developers using CSAT or NPS — vital for connecting platform improvements to developer productivity” — DORA 2025 Platform Engineering capability page
Labels	process, observability, p2-important, effort-small
Effort	Small (half-day setup, monthly practice)
Assignee	👤 Human
Prerequisites	METRICS-01, METRICS-02

---

Why DORA 2025 Specifically Requires This

The 2025 DORA research introduced a critical distinction that previous DORA reports missed: software delivery metrics (deployment frequency, lead time, etc.) tell you what is happening but not why.

The 2025 report introduces DevEx (Developer Experience) measurement alongside delivery metrics for exactly this reason. The data shows:

“The platform capability most correlated with a positive user experience is giving ‘clear feedback on the outcome of my tasks’.”

For a vibe-coding workflow, this has a direct implication: if Copilot is generating code but you don’t know whether it’s making your work feel better or worse, you are flying blind. The 2025 report also found that developer experience — specifically whether developers enjoy the workflow — is a massively underrated predictor of sustained high performance.

DORA recommends tracking DevEx via regular CSAT or NPS surveys, and connecting those scores to platform/tooling changes over time.

---

What DevEx Means for the Goald Project

For a solo or small team, a formal NPS survey is overkill. But a structured monthly 5-question self-assessment captures the same signal with negligible overhead. The five dimensions come directly from the DORA/SPACE frameworks:

Dimension	Question	Score (1–5)
Flow	How often do I reach a flow state during development sessions this month?
Feedback speed	How fast does the system (CI, Copilot, tests) tell me when something is wrong?
Cognitive load	How much mental effort does navigating the codebase and tooling require?
AI trust	How often do I accept Copilot output with confidence vs. needing to rewrite?
Tooling friction	How often does a tool, script, or process block my work?

A score < 3 on any dimension for two consecutive months is a trigger to file an improvement issue.

---

Acceptance Criteria

DevEx Tracking Template

• docs/DEVEX_LOG.md created with a table template for monthly self-assessment:

  | Month | Flow | Feedback Speed | Cognitive Load | AI Trust | Tooling Friction | Notes |
  |---|---|---|---|---|---|---|
  | 2026-03 |   |   |   |   |   |   |

• First entry completed for the current month

• docs/RUNBOOKS.md updated with “Monthly DevEx Review” procedure (5 minutes, last Friday of each month)

Clear Feedback Improvements (direct DORA 2025 recommendation)

The specific platform capability DORA found most correlated with positive DevEx is “clear feedback on the outcome of tasks.” For Goald, this means:

• CI output is human-readable: each step has a clear ✅/❌ with a one-line explanation, not just exit codes
• Test failure output includes the file, line, expected value, and actual value — not just “FAIL”
• ESLint errors in CI include the rule name and a link to the rule documentation
• All npm scripts in package.json produce a completion message: e.g., echo "✅ Lint passed" or use a script wrapper
• scripts/weekly-metrics.sh (from METRICS-02) produces a single-screen summary with ✅/⚠️/❌ status indicators

Trigger Protocol

• docs/DEVEX_LOG.md includes a “Trigger Table”:

  Score	Action
  Any dimension < 3 for 1 month	Note and monitor
  Any dimension < 3 for 2 consecutive months	File an improvement issue immediately
  Cognitive Load ≥ 4 for 1 month	Run VSM-01 review — the codebase may have accumulated complexity
  AI Trust < 3 for 1 month	Run PROCESS-02 copilot-instructions tuning immediately
  Tooling Friction ≥ 4 for 1 month	File a PLAT-01 follow-up issue

Connection to WELLBEING-01

• docs/SUSTAINABLE_PACE.md (from WELLBEING-01) updated to include the DevEx monthly score as the single leading indicator of team health

---

Files to Create / Change

• docs/DEVEX_LOG.md (create)
• docs/RUNBOOKS.md — add Monthly DevEx Review procedure
• docs/SUSTAINABLE_PACE.md — add DevEx score as leading indicator
• package.json — add completion messages to all npm scripts
• .github/workflows/ci-quality.yml — improve step output readability