ワンクリックでManusで任意のスキルを実行

meta-dream-nap

スター0

フォーク0

更新日2026年7月18日 18:18

Use for a QUICK, cheap memory consolidation of only the current directory's chain - before or right after a context compaction (the PostCompact nudge points here), when the session accumulated learning signals worth folding in NOW, or on "nap", "dream nap", "/dream-nap", "quick tidy". Minutes, not tens of minutes; for the full periodic consolidation of the whole tree use bitranox:meta-dream-tree.

インストール

Codex または Claude でインストールこの Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。

Manusで実行

ソース

bitranox

bitranox/bitranox-skills

GitHub リポジトリを開く Creator のリポジトリを見る

ダウンロード

Manusで実行

meta-dream-nap

The NAP: a light, frequent consolidation of ONLY the cwd's altitude chain (project -> ancestors -> anchor). It exists because the full dream is tree-wide and heavy; skipping consolidation at a compaction loses detail, and silently cherry-picking the full dream's steps leaves a half-consolidated store. A nap does the chain-internal basics fast and REPORTS what it deferred.

REQUIRED BACKGROUND: bitranox:meta-dream-tree -> references/dream-core.md - the shared core (mode, capture-first, backup+manifest, dedup semantics, routing prompt, verification, tiers) lives there and applies here unchanged; this file only carries the nap's SCOPE DELTA. The storage spec is bitranox:meta-self-improve -> references/memory-backend.md.

Scope (the delta)

The nap's scope is the cwd's ALTITUDE CHAIN ONLY: the project level, each ancestor level, and the anchor. SIBLING projects, sibling departments, and other trees are OUT OF SCOPE - a nap never reads or writes them (the acceptance harness asserts they stay byte-identical). Everything cross-sibling belongs to bitranox:meta-dream-tree; everything cross-tree to bitranox:meta-dream-crosstree.

When to run

Around a context compaction: the PreCompact hook salvaged candidate learnings; nap them into the store while the detail is still warm (the PostCompact nudge points here).
The session accumulated several learning signals and a full dream is not worth its cost now.
Manual: "nap", "dream nap", "/dream-nap", "quick tidy".
NOT a replacement for the full dream: the SessionStart consolidation-due nudge still means bitranox:meta-dream-tree.

Procedure (all semantics per dream-core.md)

Capture first (unconditional; the audit/salvage candidates plus this session's signals) - via bitranox:meta-self-improve, at the project level.
Back up + manifest the CHAIN's levels only (per the core).
Dedup within the chain: fold near-duplicates among the chain's entries (engine add, same slug). Do not chase sibling duplicates - out of scope, the full dream's job.
Chain-internal placement only: route entries whose EVIDENCE is already clear through the routing prompt (in the core), applying only moves whose from AND to levels are ON the chain (up or down). Low/UNSURE and anything involving a sibling stays put. Pinned entries untouched.
Prune the obvious: leaked task-state and dead-content entries AT the chain's levels (propose-first per the removal policy).
Verify + report + state the leftovers: the core's verification contract (manifest diff, reconcile TOTAL problems: 0 over the chain), then report counts AND an explicit "deferred to the full dream" list (sibling dedup, tree-wide placement, descriptor synthesis, behavioral passes, and the toolbox pass per dream-core.md - machine-global, outside the nap's chain-only budget). A nap is PARTIAL BY DESIGN and says so - never silently incomplete. Run dream_state.py done (home: <plugin>/skills/meta-dream-tree/dream_state.py, via hooks/run-python.sh - see dream-core.md "Script homes") ONLY if the nudge that triggered you asked for a nap; a consolidation-due nudge still needs the full dream.

Deliverables (a completed nap has ALL of these)

Capture ran (or a verified "nothing durable" line).
Backup + manifest of the chain's levels recorded.
Chain-internal dedup/placement/prune applied via the engine (fail-loud success lines).
Sibling branches and other trees untouched (nothing outside the chain was read or written).
Manifest diff clean; reconcile TOTAL problems: 0 over the chain.
The report ends with the explicit DEFERRED list for the next full dream.

Rationalizations (these do not fly)

Excuse	Reality
"While I'm here, that sibling duplicate is one quick merge"	Out of scope by design - the sibling snapshot assertion fails and the nap becomes a slow dream.
"No time even for a nap - skip consolidation"	The nap exists precisely for this budget; capture + chain dedup is minutes.
"I'll just cherry-pick the full dream's steps"	That is the silent-partial trap the nap replaces: run the nap, get the DEFERRED list stated.
"The nap ran, so consolidation is done"	A nap is partial by design; the consolidation-due nudge still means the full dream.

Common mistakes

Touching a sibling project/department or another tree (scope violation; harness-asserted).
Skipping the DEFERRED list (the next full dream needs to know what a nap left behind).
Running dream_state.py done after a nap that was triggered by a consolidation-due nudge (silences the nudge without the tree-wide work having happened).
Restating core semantics here instead of following dream-core.md (the contract test fails duplicated family literals).

このリポジトリの他の Skills

同じリポジトリ

compuse-bash

bitranox/bitranox-skills

Use when running bash/shell commands and interpreting their exit codes and output - pipelines, chaining a check after a mutating command, process checks with pgrep/pkill, backgrounding, waiting for an event, or when a command "failed" or its result looks ambiguous.

2026-07-200

meta-self-improve

bitranox/bitranox-skills

Use at the end of a turn that produced a learning, such as a correction from the user, a rule or preference stated ("remember this", "from now on", "always/never"), a process or tooling mistake (wrong command, a quirk of the shell/SSH/environment, a misread of stale output, over-waiting, a tool or file you missed and re-derived), a wasted build or test cycle, or any reusable discovery (a procedure, a timing, a gotcha, a flag combination, a path). A gated Stop hook nudges this automatically; also run on "self-improve", "/self-improve", "improve the harness", or "capture what we learned".

2026-07-200

coding-python-new-public-library

bitranox/bitranox-skills

Use when creating or scaffolding a NEW public bitranox Python library - a pip-installable package with a rich-click CLI, CI across Python 3.9-3.14 on Linux/macOS/Windows, and a PyPI release - or when installing or using a library built from the bitranox_template_py_lib template. Keywords - new library, scaffold, project template, rename-project, bmk, make test, make release, uv pip install.

2026-07-190

coding-python-pwshpy

bitranox/bitranox-skills

Use when issuing or running PowerShell (pwsh) commands, or writing, modifying, or debugging a PowerShell / .ps1 script - prefer pwshpy instead: a genuinely Pythonic PowerShell (typed records + a lazy pipeline over native OS bindings, never a pwsh.exe subprocess, never text parsing), as a Python library (from pwshpy import ps) or the pwshpy CLI (install/run via uv - `uv tool install pwshpy` or `uvx pwshpy`). Better error handling and reporting (real exceptions and real exit codes, nothing silently swallowed) plus clean typed JSON instead of Get-* text. Covers using pwshpy and translating cmdlets / .ps1 to it.

2026-07-190

coding-python-send-mail

bitranox/bitranox-skills

Use when sending email from Python or the shell, especially with large attachments that must not be loaded into memory, RFC 3030 BDAT/CHUNKING, STARTTLS with authentication, multi-host failover, or correct UTF-8 subject and body encoding. Prefer the `btx_lib_mail` library or its `btx-lib-mail` CLI (zero-install via `uvx btx-lib-mail send ...`) over hand-rolling `smtplib`/`email`, MIME assembly, dot-stuffing, or attachment security checks. Covers install, uvx, the library API, the CLI, streaming and BDAT, and attachment security.

2026-07-190

meta-skill-writer

bitranox/bitranox-skills

Use when creating new skills, editing existing skills, structuring SKILL.md files, writing skill frontmatter, testing skills with subagents, deploying skills, or verifying skills work before deployment

2026-07-190

name	meta-dream-nap
description	Use for a QUICK, cheap memory consolidation of only the current directory's chain - before or right after a context compaction (the PostCompact nudge points here), when the session accumulated learning signals worth folding in NOW, or on "nap", "dream nap", "/dream-nap", "quick tidy". Minutes, not tens of minutes; for the full periodic consolidation of the whole tree use bitranox:meta-dream-tree.

meta-dream-nap

Scope (the delta)

When to run

Around a context compaction: the PreCompact hook salvaged candidate learnings; nap them into the store while the detail is still warm (the PostCompact nudge points here).
The session accumulated several learning signals and a full dream is not worth its cost now.
Manual: "nap", "dream nap", "/dream-nap", "quick tidy".
NOT a replacement for the full dream: the SessionStart consolidation-due nudge still means bitranox:meta-dream-tree.

Procedure (all semantics per dream-core.md)

Capture first (unconditional; the audit/salvage candidates plus this session's signals) - via bitranox:meta-self-improve, at the project level.
Back up + manifest the CHAIN's levels only (per the core).
Dedup within the chain: fold near-duplicates among the chain's entries (engine add, same slug). Do not chase sibling duplicates - out of scope, the full dream's job.
Chain-internal placement only: route entries whose EVIDENCE is already clear through the routing prompt (in the core), applying only moves whose from AND to levels are ON the chain (up or down). Low/UNSURE and anything involving a sibling stays put. Pinned entries untouched.
Prune the obvious: leaked task-state and dead-content entries AT the chain's levels (propose-first per the removal policy).
Verify + report + state the leftovers: the core's verification contract (manifest diff, reconcile TOTAL problems: 0 over the chain), then report counts AND an explicit "deferred to the full dream" list (sibling dedup, tree-wide placement, descriptor synthesis, behavioral passes, and the toolbox pass per dream-core.md - machine-global, outside the nap's chain-only budget). A nap is PARTIAL BY DESIGN and says so - never silently incomplete. Run dream_state.py done (home: <plugin>/skills/meta-dream-tree/dream_state.py, via hooks/run-python.sh - see dream-core.md "Script homes") ONLY if the nudge that triggered you asked for a nap; a consolidation-due nudge still needs the full dream.

Deliverables (a completed nap has ALL of these)

Capture ran (or a verified "nothing durable" line).
Backup + manifest of the chain's levels recorded.
Chain-internal dedup/placement/prune applied via the engine (fail-loud success lines).
Sibling branches and other trees untouched (nothing outside the chain was read or written).
Manifest diff clean; reconcile TOTAL problems: 0 over the chain.
The report ends with the explicit DEFERRED list for the next full dream.

Rationalizations (these do not fly)

Excuse	Reality
"While I'm here, that sibling duplicate is one quick merge"	Out of scope by design - the sibling snapshot assertion fails and the nap becomes a slow dream.
"No time even for a nap - skip consolidation"	The nap exists precisely for this budget; capture + chain dedup is minutes.
"I'll just cherry-pick the full dream's steps"	That is the silent-partial trap the nap replaces: run the nap, get the DEFERRED list stated.
"The nap ran, so consolidation is done"	A nap is partial by design; the consolidation-due nudge still means the full dream.

Common mistakes

Touching a sibling project/department or another tree (scope violation; harness-asserted).
Skipping the DEFERRED list (the next full dream needs to know what a nap left behind).
Running dream_state.py done after a nap that was triggered by a consolidation-due nudge (silences the nudge without the tree-wide work having happened).
Restating core semantics here instead of following dream-core.md (the contract test fails duplicated family literals).