Skip to content
Back to skills

Cocolean

ASecurity

CocoLean shortcut debt ledger — harvests cocoplus: comment markers across the CocoPod and produces a ranked deferral report sorted by ceiling imminence and age.

  • 724 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added September 5, 2026
ai-agentsjavascriptjavashellsqlnodegitperformance

Security analysis

A100/100

Pro scans all 6 files and shows the line behind each finding

Scanned September 5, 2026

npx -y skills add Snowflake-Labs/cocoplus --skill cocolean --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Cocolean?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Cocolean
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/snowflake-labs-cocolean-cocoplus/badge)](https://www.skillsdirectory.com/skills/snowflake-labs-cocolean-cocoplus)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: lean-debt
description: CocoLean shortcut debt ledger — harvests cocoplus: comment markers across the CocoPod and produces a ranked deferral report sorted by ceiling imminence and age.
version: "1.0.0"
author: sgsshankar
tags:
  - cocolean
  - technical-debt
  - delivery-discipline
user-invocable: true
blocking: true
---

## Objective

You are executing `$lean debt` — the CocoLean shortcut debt ledger. Harvest all `cocoplus:` comment markers across the CocoPod, assess ceiling imminence, and produce a ranked deferral report. Run this before sprint planning and before any release gate.

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

## Step 1 — Run Deterministic Marker Harvest

```
invoke cocolean/lean-debt-engine
```

Run the command directly with Node. Do not wrap it in PowerShell redirection such as `2>&1`; stderr handling differs across shells and can corrupt the JSON output.

The script:
1. Scans all `.sql`, `.js`, `.md`, `.yaml`, `.json` files in the CocoPod for `cocoplus:` comment markers
2. Extracts the three required fields from each marker: **what was simplified**, **ceiling condition**, **upgrade trigger**
3. Calculates ceiling imminence from current project metrics (from `dora-snapshot.json`, `flow.json`, `meta.json`)
4. Outputs a JSON array of debt records sorted by `(ceiling_imminence × days_since_annotation)`

Read the JSON output.

## Step 2 — The cocoplus: Comment Protocol

A valid `cocoplus:` marker looks like this:

```sql
-- cocoplus: simplified=single-warehouse routing; ceiling=will fail when concurrent agent count >20; trigger=when $build spawns >20 agents in one run
```

```javascript
// cocoplus: simplified=linear scan for seed matching; ceiling=scan time exceeds 5s above 500 seeds; trigger=when $seed list reports >500 entries
```

Three required fields (pipe-delimited or semicolon-delimited):
- `simplified=` — one sentence describing the shortcut taken and why
- `ceiling=` — the condition under which this shortcut fails or requires rework
- `trigger=` — the specific event that should prompt addressing the ceiling

Markers missing any required field are reported as malformed but not blocked.

## Step 3 — Produce Ranked Deferral Report

Write the report to `.cocoplus/lifecycle/lean-debt.md` and display a summary. The summary must include a compact dashboard table grouped by imminence and a trigger surface listing the next concrete events that would force action.

```markdown
# CocoLean Debt Ledger
Generated: [ISO 8601 timestamp]
CocoPod: [project name from meta.json]

## Summary
Total markers: [N]
High imminence (address this sprint): [N]
Medium imminence (address this quarter): [N]
Low imminence (monitor): [N]
Malformed markers (missing fields): [N]

---

## High Imminence

### [1] scripts/the SessionStart hook:47
**Simplified:** Single-threaded seed evaluation — evaluates all seeds synchronously on session start
**Ceiling:** Session start latency exceeds 500ms when seed count exceeds 200
**Trigger:** When `$seed list` reports more than 200 entries
**Age:** 43 days since annotation
**Ceiling Imminence:** HIGH — current seed count: 187 (94% of ceiling)
**Recommended Action:** Migrate seed evaluation to async Tier 2 pattern before next sprint

---

## Medium Imminence

### [2] skills/cocoflow/flow-run.skill.md:23
**Simplified:** In-memory stage dependency graph — rebuilt from flow.json on every $flow run
**Ceiling:** Rebuild time exceeds 2s for pipelines with >50 stages
**Trigger:** When any flow.json exceeds 50 stages
**Age:** 12 days since annotation
**Ceiling Imminence:** MEDIUM — current max stage count: 18 (36% of ceiling)

---

## Low Imminence

[remaining entries]

---

## Malformed Markers

[file:line] — missing field: [field name]

---

## Trigger Surface

| Trigger | Current Signal | Debt Items Activated |
|---------|----------------|----------------------|
| [specific event] | [current metric or unknown] | [file:line list] |

## Integration Notes

- Link CocoReview or SecondEye findings that mention the same file:line as a debt marker.
- When a review finding is intentionally deferred, add or update a `cocoplus:` marker instead of leaving the deferral only in prose.
```

## Step 4 — Commit the Report

Create git commit:
```
chore(lean): update shortcut debt ledger [timestamp]
```

## Step 5 — LLM Summary Layer

After the deterministic harvest, use Haiku to generate a one-paragraph narrative assessment:
- Which debt items represent the most systemic risk?
- Are there patterns in the type of shortcuts taken (all performance shortcuts? all scale shortcuts?)
- What does the age distribution suggest about the team's debt repayment cadence?

Append this narrative to the end of `lean-debt.md` under `## Narrative Assessment`.

## Exit Criteria

- `cocolean/lean-debt-engine` executed across full CocoPod
- All `cocoplus:` markers harvested and classified by imminence
- Report ranked by `(ceiling_imminence × days_since_annotation)`
- Report written to `.cocoplus/lifecycle/lean-debt.md` and committed
- Malformed markers surfaced without blocking
- Haiku narrative assessment appended

## Anti-Rationalization

| Temptation | Why Wrong |
|------------|-----------|
| Condemn shortcuts with HIGH imminence as mistakes | The ledger records reasonable decisions at the time — the shortcut was right when made; only the ceiling date has changed |
| Skip the report if the debt count is low | Zero markers may mean no shortcuts were taken, or may mean `cocoplus:` protocol was not followed — surface which it is |
| Omit the LLM narrative because the deterministic output is sufficient | The deterministic output ranks by formula; the narrative identifies systemic patterns the formula cannot see |
| Hide the trigger surface inside individual findings | Sprint planning needs a compact list of upcoming events that activate deferred work |

Files in this skill

  • lean-debt-engine.skill.md1.5 KB
  • lean-debt.skill.md5.8 KB
  • lean-review-engine.skill.md1.5 KB
  • lean-review.skill.md5.6 KB
  • lean.skill.md5.4 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…