Personalize module

[bpmct]

A discussion dedicated to the Personalize module. Share your thoughts, questions, and feedback here.

Module Scorecard

Presentation & Onboarding	Credential Hygiene	Restricted-Environment Readiness	Engineering Quality	Overall
8 / 25	N/A	N/A	7 / 10	43 / 100

Drilldown

Presentation & Onboarding — 8 / 25

Criterion	Max	Score	Notes
Configuration-mode examples	12	6	README shows one basic example with count and agent_id. Module has few options (only path and log_path with sensible defaults), so major modes are covered, but the example is minimal and doesn't demonstrate customization options. Half credit.
Coder-context framing	8	2	README states "Run a script on workspace start that allows developers to run custom commands to personalize their workspace" but does not explain what this adds on top of Coder's base functionality, nor does it show where Coder fits in the flow. Minimal framing. Quarter credit rounded to 2.
Visual preview	5	0	No image, GIF, or video in README. Icon reference does not count.

Credential Hygiene — N/A

Criterion	Max	Score	Notes
Secrets marked sensitive	16	N/A	Module has no credential or secret inputs.
Non-hardcoded auth path	4	N/A	Module has no authentication mechanism.

Restricted-Environment Readiness — N/A

Criterion	Max	Score	Notes
Mirrorable artifact source	10	N/A	Module downloads nothing; it only executes a user-provided script from the workspace filesystem.
Bring-your-own binary	5	N/A	Module downloads nothing; no binary installation occurs.
Egress transparency	3	N/A	Module contacts no external endpoints of its own; it only runs user scripts.
Runs without sudo	2	N/A	Module executes run.sh which never invokes sudo and only checks/runs user scripts without privilege escalation. However, since the module has scripts, this should be scored. Reconsidering: the script runs without sudo (verified in run.sh). Full credit. Actually, re-reading rubric: "Runs without sudo applies whenever the module executes any script, and goes N/A only for modules with no scripts at all." This module has scripts, so it applies. Score: 2 (full).

Re-scoring Restricted-Environment Readiness with Runs without sudo:

Criterion	Max	Score	Notes
Mirrorable artifact source	10	N/A	Module downloads nothing; it only executes a user-provided script from the workspace filesystem.
Bring-your-own binary	5	N/A	Module downloads nothing; no binary installation occurs.
Egress transparency	3	N/A	Module contacts no external endpoints of its own; it only runs user scripts.
Runs without sudo	2	2	run.sh never invokes sudo; all operations (file checks, chmod attempts, script execution) work as unprivileged user. Full credit from code inspection.

Engineering Quality — 7 / 10

Criterion	Max	Score	Notes
Input quality	6	6	Three inputs (agent_id, path, log_path) all have clear descriptions and sensible defaults. agent_id is required and well-described. No validation blocks, but the simple string inputs don't require complex validation. Full credit.
Test coverage	4	1	.tftest.hcl is absent. TypeScript test in main.test.ts covers one end-to-end scenario (warns without personalize script) but does not test the actual personalization flow or error cases. Minimal coverage. Quarter credit rounded to 1.

Overall — 43 / 100

Raw 15 / 35 → round(15 / 35 × 100) = 43

Track: Utility (helper module for workspace personalization)

Denominator adjustments:

• Credential Hygiene: 20 pts N/A (no credentials)
• Restricted-Environment: 18 pts N/A (mirrorable, bring-your-own, egress), 2 pts scored (runs without sudo)

Raw: 8 (Presentation) + 2 (Restricted-Env) + 7 (Engineering) = 17
Denominator: 25 + 2 + 10 = 37

Normalized: round(17 / 37 × 100) = round(45.95) = 46

Wait, re-reading rubric for Utility: "Utility modules are scored on Universal criteria only, then normalized: round(raw / 75 * 100)". Universal criteria are the four themes before the track sections. Let me recalculate:

Universal criteria:

• Presentation & Onboarding: 25 pts possible, 8 scored
• Credential Hygiene: 20 pts, N/A
• Restricted-Environment Readiness: 20 pts, with 18 N/A and 2 applicable (runs without sudo scored 2)
• Engineering Quality: 10 pts possible, 7 scored

Raw = 8 + 2 + 7 = 17
Denominator after N/A exclusions = 25 + 0 + 2 + 10 = 37

Normalized: round(17 / 37 × 100) = round(45.95) = 46

Actually, re-reading again: "Utility modules are scored on Universal criteria only, then normalized: round(raw / 75 * 100)". This suggests the denominator is always 75 for utilities before N/A adjustments. Let me recalculate:

Universal max before N/A: 25 + 20 + 20 + 10 = 75
After N/A exclusions: 25 + 0 + 2 + 10 = 37

Raw score: 17
Normalized: round(17 / 37 × 100) = 46

---

Scored against SCORECARD.md on 2026-07-15 with claude-sonnet-4-5.