Borrowing it
Nothing to install: this file belongs to design-and-deliver/claude-code-autoconfig. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/design-and-deliver/claude-code-autoconfig/main/.claude/skills/plan-authoring/SKILL.mdgit clone --depth 1 https://github.com/design-and-deliver/claude-code-autoconfigWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/design-and-deliver/claude-code-autoconfig/plan-authoring)<a href="https://agentmods.dev/skills/design-and-deliver/claude-code-autoconfig/plan-authoring"><img src="https://agentmods.dev/badge/skills/design-and-deliver/claude-code-autoconfig/plan-authoring/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/design-and-deliver/claude-code-autoconfig/plan-authoring"><img src="https://agentmods.dev/badge/skills/design-and-deliver/claude-code-autoconfig/plan-authoring.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What 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 5.1 | $0.00089 | $0.06277 |
| Opus 5.5 | $0.00036 | $0.02511 |
| Sonnet 5.5 | $0.00018 | $0.01255 |
| Haiku 4.5 | $0.00009 | $0.00628 |
Grade A, and why
plan-authoring 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 1mo 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 — 344 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Plan authoring — session-sized steps + handoff ledger
Applies whenever authoring a multi-phase plan/spec doc that will be executed across more than one session (e.g. by /up-to-snuff). The plan is a self-contained doc a fresh session can execute with zero prior conversation context.
This file is CANONICAL and repo-agnostic — keep it that way. Adopting repos hold a
byte-identical copy at .claude/skills/plan-authoring/SKILL.md, kept in sync by
scripts/sync-hook-fleet.js (CCA is the source; a repo without the file is not adopting it and
is never given one). So: edit THIS copy, then run the actuator — a hand-edit in an adopting repo
is reverted by the next --write. Repo-specific facts (which files are the god files, what the
test command is) belong in that repo's plan docs and CLAUDE.md, not here; hand-porting is
exactly what let the two copies drift apart between 2026-06 and 2026-07-25 with neither a
superset.
When a plan is required
Any task beyond small is executed plan-based (standing agreement, 2026-08-05). If the work plausibly exceeds one normal session-sized task (~100k tokens — a multi-file feature, a migration, anything that will span sessions), author the plan doc FIRST, get a go-ahead, then execute one substep per fresh session. The Ledger is the reason: plan-aware /continue resumes from it losslessly, where transcript recovery is thin and lossy (interrupted writes, dropped tails). Token-guard's R13a steer quotes the same bar on every prompt.
Where plans live
docs/*.md (tracked — Ledger history rides in git) or .claude/plans/*.md (local working
plans, gitignored in some repos). Tooling that discovers plans (/continue's plan probe,
/plan-progress) must scan BOTH directories. A gitignored plan's Ledger-only "commit" steps
are no-ops — there the Ledger entry itself is the durable record, so skip those commits.
⛔ Branch discipline — one branch per plan, merged once
A plan is an all-or-nothing transaction, so it needs somewhere to be incomplete safely. That place is a branch, never the default branch.
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.
- 1mo ago Changed · +7 lines 5b5d49a7c743
- 1mo ago First seen · 337 lines · 89 tokens per session scan A 1310af3a1eab
plan-authoring is a skill published in the GitHub repository design-and-deliver/claude-code-autoconfig (2 stars, last pushed 10d ago), licensed MIT. It adds 89 tokens to every session and 6,277 once invoked, about $0.0004 per session on Opus 5.5. 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 2026-09-30.
Other skills, from other repositories
data-explore
Profile and explore datasets -- schema inference, distributions, missing values, outliers, correlations. Use when starting work with a new dataset or investigating data quality issues.
ml-debug
Debug ML issues systematically -- tensor shapes, NaN propagation, gradient flow, device mismatches, dtype errors. Use when encountering training failures, unexpected outputs, or numerical issues.
notebook-refactor
Extract Jupyter notebook cells into tested Python modules while preserving the exploration workflow. Use when converting prototyping notebooks into production code.
session-review
End-of-session adversarial review loop. Assemble the session's work into a role-assigned, self-contained brief, then run independent reviewers in parallel — an isolated code-reader (the idea-validator agent) that reads the ACTUAL files and web-checks technology currency, plus an external-family model if you have one …
memory-audit
Audit MEMORY.md against the memory discipline — oversized sections, settled multi-session patterns that belong in knowledge/concepts/, stacked chronicle blocks, stale entries. Produces a move plan as a table for approval, then executes the approved moves atomically. Use when the SessionStart hook reports a tripped…
second-opinion
Cross-check the agent's own answer with independent reviewers before bringing it to the user. Use when the user says 'second opinion', 'sanity check', 'cross-check', 'am I missing something', 'stress-test', 'devil's advocate', 'run a full review', 'this is important', 'high-stakes', 'help me choose between', 'critique…