Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/superplan-md/superplan-plugin/superplan-reviewnpx skills add superplan-md/superplan-plugin --skill superplan-reviewgit clone --depth 1 https://github.com/superplan-md/superplan-pluginWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00027 | $0.02038 |
| Opus 5 | $0.00014 | $0.01019 |
| Sonnet 5 | $0.00005 | $0.00408 |
| Haiku 4.5 | $0.00003 | $0.00204 |
Grade A, and why
superplan-review scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 270 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Review Task Against AC
Overview
Decide whether a task is actually complete against its contract.
This is the Superplan-specific completion gate. It must judge completion against the task contract, acceptance criteria, and fresh evidence rather than vibes or implementation plausibility.
Hard Gate
Do not bless completion when the evidence is stale, partial, or weaker than the acceptance criteria require.
Trigger
Use when:
- a task is claimed ready for completion
- downstream work may unblock if the task is truly done
- verification has run and the system must decide whether the results are still valid
- subagents have produced evidence that appears sufficient
Stay Out
Do not review when:
- the task is clearly mid-execution
- key evidence is still missing
- the task contract is too broken to review cleanly
- the verification basis is obviously stale and must be rerun first
- structural trajectory change invalidated the review target
Inputs And Assumptions
Inputs:
- task contract
- acceptance criteria
- implementation evidence
- verification outputs
- outputs from trusted workspace-native harnesses or QA flows
- available repo-native scripts, custom skills, browser flows, QA routines, and contract checks
- runtime notes and recent trajectory changes
- subagent outputs and ownership boundaries where relevant
- chosen verification plan when available
- recent decision log entries affecting contract interpretation
- known gotchas that may invalidate naive review
Assumptions:
- completion must be judged against the task contract
- evidence can become stale after material implementation or trajectory change
- not all evidence is equal
- the strongest honest proof should come from an existing trusted workspace harness when possible
- if no existing harness proves the AC directly, review should require the smallest credible verification loop that does
- static analysis is a constrained fallback, not a default completion oracle
Verification Acquisition Ladder
What ships with it
10 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- evals/01-workspace-harness-first.md 556 B
- evals/02-derive-targeted-verification.md 556 B
- evals/03-static-analysis-fallback.md 543 B
- evals/04-stale-browser-subagent-proof.md 414 B
- evals/05-contradictory-contract.md 419 B
- evals/README.md 1.8 KB
- references/evidence-validity.md 1.6 KB
- references/gotchas.md 896 B
- references/stale-verification.md 424 B
- references/verification-acquisition-ladder.md 2.2 KB
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 270 lines · 27 tokens per session scan A 86854e5e6907
superplan-review is a skill published in the GitHub repository superplan-md/superplan-plugin (72 stars, last pushed 3mo ago), licensed Apache-2.0. It adds 27 tokens to every session and 2,038 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
improve
Autonomous quality improvement loop. Scores a target against a rubric, selects the highest-leverage axis, attacks it, verifies, documents, and loops. No pre-planning between iterations — each loop re-scores from scratch.
setup
First-run experience for the harness. Three modes: Recommended (guided, 3 min), Full Tour (guided + skill walkthrough, 8 min), and Express (zero questions, 30 sec). Installs hooks first, detects stack, configures harness.json, runs a live demo on real code, and prints a reference card.
evolve
Research-driven multi-cycle improvement director. Forms causal hypotheses about why scores are low, validates them with scout agents before attacking, dispatches axis-parallel fleet attacks, extracts transferable patterns, and runs indefinitely within a budget envelope. Accumulates a persistent belief model and…
research
Focused research investigations. Converts questions into structured findings with confidence levels and source citations. Single agent by default; with --parallel (or when the question decomposes into 3+ independent angles) it spawns scout agents whose findings are compressed into a unified brief. Does not make…
ascii-diagram
Generate perfectly aligned ASCII diagrams — architecture, flow, sequence, box-and-arrow. Uses a programmatic character-grid approach so alignment is guaranteed by math, not token prediction. Includes post-render verification.
triage
GitHub issue and PR investigator. Pulls open issues/PRs, classifies them, searches the codebase for root cause or reviews contributed code, proposes fixes with file:line references, and optionally implements fixes. Use for investigating GitHub issues and reviewing PRs; do NOT use for general code review unrelated to…