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 commands/w00fx/spec-anchored-agentic-development/explaingit clone --depth 1 https://github.com/w00fx/spec-anchored-agentic-developmentWhat 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 | $0.00046 | $0.01945 |
| Opus 5 | $0.00023 | $0.00972 |
| Sonnet 5 | $0.00009 | $0.00389 |
| Haiku 4.5 | $0.00005 | $0.00194 |
Grade A, and why
explain 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 — 141 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Generate a complete reference walkthrough of the changes in the argument — the work just completed in this session by default, or a diff/branch/PR if given. The goal is a document the human can read end to end and come away knowing exactly what was built and why, with nothing material left out.
Gather the full context first
Do not walk the diff blind. Read, and use as the backbone of the explanation:
- The diff of all changes (the what).
- The capability spec and the approved plan (protocol Phase 3) (the intent — what was asked and which approach was committed).
- The decisions captured in Phases 1.5 and 2, including the tradeoffs (the why, which the code does not contain).
- The lessons and any low-confidence decisions from Phase 6.
- The constitution (for the domain rules the change touches).
The intent artifacts are what make this more than a code re-narration: explain the change against what it was supposed to do, not in a vacuum.
When the target is work from another session (a diff, branch, or PR), the
intent artifacts aren't in this conversation: pull the plan from the PR
description (both implement skills put it there) or the run log under
.claude/logs/. If an intent artifact is genuinely unavailable, say which
dimensions lack it — do not reconstruct intent from the code and present it
as the plan.
The bar: complete in coverage, not in granularity
Cover every dimension below — leave none out. But explain at the level of why and how it connects, not line by line. If a paragraph could be replaced by reading the code, cut it; spend the words on what the code does not say (the reason, the constraint, the cross-file flow, the edge case). "Exhaustive" means no dimension missing, not every line narrated. There is no length limit — let completeness set the length. And when behavior resists prose — a scheduler, a state machine, a data flow you can only feel by stepping through it — offer a throwaway comprehension tool alongside the document (a custom debugger, a step-through UI, a time-scrubber): agents can write code that helps humans understand code; keep the understanding, delete the tool. And for behaviors with state or permissions, cover the behavior-unit checklist: trigger → rule → state change → execution path (including failure) → edge cases, each step with its evidence link — prose explains; facts anchor.
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 · 141 lines · 46 tokens per session scan A 2be29632b4ef
explain is a command published in the GitHub repository w00fx/spec-anchored-agentic-development (5 stars, last pushed 7d ago), licensed MIT. It adds 46 tokens to every session and 1,945 once invoked, about $0.0002 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 2026-08-31.
Other commands, from other repositories
quiz-gate
The ★-tap NUDGE before merging a significant+worthy gate PR: a 5-question quiz built from the ACTUAL diff+tests, routed through deep-understand's mastery gate. Three logged responses (engage/defer/wave); advisory, never must-pass. Gates the human's attention, not the merge.
retrospective
Perform a post-implementation retrospective analysis measuring spec adherence, implementation deviations, and lessons learned.
auto-optimize
Self-improving agent optimization — generates challenger variants of any agent/command, benchmarks against baseline, promotes winners, logs learnings to instincts. Inspired by Karpathy's autoresearch pattern.
learn-mode
Interactive code tutor — breaks down codebase logic, explains complexities, translates technical concepts for the user. Ideal after /btw commands. Teaches the WHY behind the code, not just the WHAT.
session-validate
End-of-session self-training — captures session metrics, extracts patterns via metaclaw-learner, updates instincts, and generates optimization hypotheses for the next run.
sw-learn
Post-ship capture of patterns and learnings.