atomantic/slashdo — Repo Appraisal

Overview

What it is — slashdo (npm package slash-do) is a curated set of ~19 /do:* slash commands that install into multiple AI coding assistant environments (Claude Code, OpenCode, Antigravity CLI, Codex, Grok Build).
Problem — It replaces ad-hoc, per-project prompting for common dev workflows — committing/PR'ing, code review, dependency audits, and task planning — with a single reusable command set.
Who it's for — Developers who use one or more AI coding CLIs and want standardized, high-rigor workflows (PR creation, multi-agent review, auto-merge, dependency pruning) without rewriting prompts each time.
Notable — Explicitly prioritizes thorough software engineering over token conservation, ships a multi-agent review/enhancement pipeline (--review-with, --enhance-with across codex/claude/grok/agy), and auto-detects installed environments for a single-command setup.

Verdict

	Rating	Summary
Quality	solid (16/24)	A well-documented, actively-shipping CLI held back by near-zero stars/forks and untyped JS with no visible CI.
PAI Relevance	integrate (0.63)	Directly applicable multi-agent review/plan-task patterns fill a real gap in PAI's Development capabilities despite a Node/npm integration gap.

Quality Assessment

16/24 — maintained / adequately-documented / early-or-minimal

Health: 6/8 (maintained)

Failed:

H6: FAIL — Open Issues: 0, probe requires >0 AND <100 as a triage signal
H8: FAIL — README badges shown are npm version, environments, commands, license — no CI badge or .github/workflows/ reference

Passed:

H1: PASS — Latest Release v3.24.0 exists
H2: PASS — Released 2026-07-12, within 12 months of appraisal
H3: PASS — Last commit 2026-07-19, well within 6 months
H4: PASS — Last commit is 1 day before the appraisal date
H5: PASS — Archived: false
H7: PASS — MIT license declared

Documentation: 6/8 (adequately-documented)

Failed:

D7: FAIL — No external docs site or wiki link visible in the README excerpt
D8: FAIL — No Limitations/Caveats/Known Issues section present

Passed:

D1: PASS — Full README content present
D2: PASS — README far exceeds 1000 bytes (8KB+ excerpt provided)
D3: PASS — "Quick Start" gives npx and curl install instructions
D4: PASS — "Workflows" section shows runnable code blocks for /do:pr, /do:plan-task, /do:next
D5: PASS — "## Commands" table plus referenced "Configuration (do:config)" section
D6: PASS — Opening tagline "Curated slash commands for AI coding assistants... One install. Multiple environments. All the workflows." states purpose within first 500 characters

Engineering Signals: 4/8 (early-or-minimal)

Failed:

E1: FAIL — Primary language is JavaScript, not a typed language
E5: FAIL — 2 stars, well below the 50-star adoption threshold
E6: FAIL — Created 2026-03-03 (~4.6 months old) with 2 stars ≈ 0.4 stars/month, below the 2/month bar
E7: FAIL — 2 forks, below the 5-fork threshold

Passed:

E2: PASS — package.json manifest present with full metadata
E3: PASS — Manifest shows no "dependencies" key (minimal/zero runtime deps), well under the 15-dep CLI threshold
E4: PASS — "test": "node --test test/*.test.js" script plus dedicated test/ directory
E8: PASS — Descriptions are specific and meaningful ("Curated slash commands for AI coding assistants — Claude Code, OpenCode, Antigravity CLI, Codex, and Grok Build")

PAI Relevance

Dimension	Score	Assessment
Harvest Value	2	The `--review-with`/`--enhance-with` sequential multi-agent pipeline and PR auto-merge gate design are directly transferable patterns for PAI's Loop/Optimize/Evals iterative agent-cycle skills.
Integration Readiness	1	Node/npm CLI (`engines: node>=18`) rather than Bun+TypeScript-native; would need an adapter or subprocess wrapper to slot into PAI's CLI-first Bun stack.
Overlap Risk	1	Partial overlap with PAI's Development category (ISA, Loop, Optimize, Agents/Delegation), which already handles iterative agent-driven improvement and agent composition, though slashdo's specific PR/dependency-audit/plan-task surface isn't duplicated.
Gap Fill	1	PAI's manifest has no dedicated git/PR-workflow command bundle (commit→push→PR→review→merge, dependency pruning, task-to-issue drafting); slashdo's `/do:*` set addresses that specific automation niche.

Composite: 0.63

What Next

Conservancy editorial loop: Run npx slash-do in the Conservancy repo root to auto-install /do:* commands into Claude Code — swap the current manual commit/PR steps in the autonomous editorial pipeline for /do:commit and /do:pr, giving the loop a standardized, high-rigor path to open PRs without hand-written prompts each cycle.
Fabric Recommender (fab): Install slash-do's /do:review and /do:audit-deps commands into the fab CLI's dev environment — replaces ad-hoc dependency-pruning prompts with the built-in dependency audit command, useful now that fab is at V2.1 and accumulating dependencies across iterations.
Cross-project dev workflow: Add slash-do's multi-agent --review-with pipeline (codex/claude/grok) as the standard PR-review step for both Conservancy and fab — gives both active projects a consistent second-opinion review gate before merge, instead of relying on a single model's judgment.

Landscape Position

Category: Developer Tools & CLIs

In this category: cheat--cheat, casey--just, mvanhorn--cli-printing-press, openclaw--gogcli, simonw--showboat, gastownhall--beads, ohmyzsh--ohmyzsh, tirth8205--code-review-graph, tw93--Mole, mvanhorn--printing-press-library, mattpocock--skills

Standing: The newest and smallest-traction entrant in a crowded, well-adopted category — functionally closest to mattpocock--skills (agent workflow presets) and gastownhall--beads (coding-agent memory), but at 2 stars it lags every peer by orders of magnitude despite comparable engineering ambition.

Evidence Base

Density: 8/10 — Available: README (partial, ~8KB), package.json dependency manifest, GitHub repo metadata (stars, forks, open issues, release, commit timestamps). Missing: full README beyond the excerpt, CI/workflow config, test file contents, contribution/limitations documentation.

Notes

Very young project (created 2026-03-03) but with a rapid release cadence reaching v3.24.0 in under five months, suggesting active iteration despite minimal community traction. Conceptually overlaps with the "Agent Development Methods" cluster (mattpocock--skills, bmad-code-org--BMAD-METHOD) rather than being a pure CLI utility — worth watching for star growth as a signal of real-world adoption before considering deeper integration.