rewrite
Rewrite episodes impacted by design changes, then verify and prepare for re-polish.
Skill by Anubis44197
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/anubis44197-rewrite
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: design
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/rewrite
What's inside
- Skill name
rewrite- Description
- Rewrite episodes impacted by design changes, then verify and prepare for re-polish.
- License
- Apache-2.0
- Instructions
- 96 lines in SKILL.md · view on GitHub
Agent Skills spec issues
- files: referenced but missing: references/pii-redaction-policy.md, references/workdir-isolation-policy.md, references/error-code-glossary.md
Files (6)
- SKILL.md
- character-dialogue-dna.md · references/
- character-embodiment-rubric.md · references/
- design-document-map.md · references/
- rewrite-quality-criteria.md · references/
- rewrite-report-unified-schema.md · references/
SKILL.md (first lines)
# Rewrite Skill
## Purpose
Apply structural rewrites after design changes.
## Pipeline
1. Divergence analysis (`revision-analyst` + `character-sculptor`)
2. Rewrite execution (`episode-rewriter`)
3. Turkish language and book-mode polish (`tdk-polisher`, REWRITE mode) [mandatory]
4. Book layout normalization (`tdk-layout-agent`) [mandatory when `book_mode.enabled=true`]
5. Verification (`quality-verifier`, REWRITE mode)
6. Retry loop (bounded)
7. Mark rewritten episodes for re-polish
## Config Source
- `novel-config.md`
- `revision/_state/design-hashes.json` is the approved design baseline.
- `revision/_state/create-plan.json` is the chapter production board and must be updated when rewritten chapters change status.
## Design Change Impact Gate
Before rewriting story text, compare the approved design baseline against the current design/state files.
If any approved design source changed, emit `revision/_workspace/rewrite-impact-report.json` with:
- `run_id`
- `changed_design_sources`
- `affected_chapters`
- `impact_level` (`CRITICAL` | `MAJOR` | `MINOR`)
- `required_actions`
- `verdict` (`REWRITE_REQUIRED` | `NO_REWRITE_REQUIRED`)
No rewritten chapter may be accepted without this report when design drift exists.
## Language Policy
- Chapter/story content language must be Turkish.
- Skill/agent contracts and tooling instructions remain English.
- Preserve valid UTF-8 Turkish characters; mojibake or unexplained non-Turkish script usage must be reported as a quality issue.
## Security and Privacy Policy
- Apply PII redaction policy to reports/logs: `references/pii-redaction-policy.md`.
- Respect WORK_DIR isolation boundaries: `references/workdir-isolation-policy.md`.
- If boundary violation is detected, block step with `E_WORKDIR_BOUNDARY`.
## Runtime Metadata Contract
- Every rewrite run must have a `run_id` in format: `RUN-YYYYMMDD-HHMMSS-XXXX`.
- Every pipeline step must have a `step_id` in format: `rewrite-<step-number>`.
- All generated reports must include `run_id` and `step_id` in header metadata.
- Rewrite flow must update `{WORK_DIR}/_workspace/run-summary.json` after each step.
- For blocked/failed steps, `error_code` must come from `references/error-code-glossary.md`.
- Rewrite flow must emit `{WORK_DIR}/_workspace/pipeline-metrics.json`.
- Rewrite flow must emit `{WORK_DIR}/_workspace/pipeline-bottleneck-report.md`.
## Model Routing Policy
- Model selection must follow:
- `skills/polish/references/model-capability-matrix.md`
- `skills/polish/references/model-fallback-timeout-policy.md`
- A/B prompt experiments must follow:
- `skills/polish/references/prompt-ab-experiment-spec.md`
## Model Adapter Contract
- Shared task envelope must follow:
- `skills/polish/references/shared-task-schema.md`
- Agent/skill mapping must follow:
- `skills/polish/references/agent-skill-schema-mapping.md`
- Adapter behavior must follow:
- `skills/polish/references/adapter-claude-codex.md`
- `skills/polish/references/adapter-generic-ide-model.md`
- Verdict/report format must follow:
- `skills/polish/references/verdict-report-standard.md`
- `skills/polish/references/workflow-report-json-schema.md`
- Inter-agent handoff payload must follow:
- `skills/polish/references/handoff-contract.md`
## Final Episode Writeback Rule (Mandatory)
- If `book_mode.enabled=true`, canonical final text source is:
- `{WORK_DIR}/_workspace/09_tdk-layout_bookmode_EP{NNN}.md`
- If `book_mode.enabled=false`, canonical final text source is:
- `{WORK_DIR}/_workspace/08_tdk-polisher_polished_EP{NNN}.md`
- Orchestrator must write canonical text back to:
- `episode/epNNN.md`
## Mandatory Artifact Gates
- Do not run `quality-verifier` before `08_tdk-polisher_issues_EP{NNN}.json` and `08_tdk-polisher_report_EP{NNN}.md` exist.
- If `book_mode.enabled=true`, do not run `quality-verifier` before `09_tdk-layout_issues_EP{NNN}.json` and `09_tdk-layout_report_EP{NNN}.md` exist.
- Do not accept `PASS` if the rewritten chapter word count is outside the `min_words` / `max_words` range in `revision/_state/create-plan.json`.
- Do not accept `PASS` if `revision/_workspace/rewrite-impact-report.json` is required but missing.
- If any mandatory artifact is missing, stop with explicit artifact-missing error.
## Outputs
- rewritten episode files
- mandatory TDK polisher outputs (`08_tdk-polisher_*`)
- mandatory layout outputs when book mode is enabled (`09_tdk-layout_*`)
- rewrite reports/logs/plans
- unified rewrite report schema:
- `skills/rewrite/references/rewrite-report-unified-schema.md`From Anubis44197/kit_hub, Apache-2.0.
More from Anubis44197/kit_hub
- 0.523score
polish
Anubis44197
Polish chapters through book-structure diagnostics, editorial correction, TDK checks, and layout review.
SkillBFreshness: BClaude Code +9 - 0.522score
create
Anubis44197
Create book chapters sequentially using mapped design docs, guard rails, continuity tracking, and automated quality gates.
SkillBFreshness: BClaude Code +9 - 0.519score
export-word
Anubis44197
Export a complete validated book package to DOCX using approval, front matter, cover brief, and book-mode style gates.
SkillBFreshness: BClaude Code +9 - 0.518score
design-small
Anubis44197
Produce detailed chapter-range design using big-design outputs as prerequisites.
SkillBFreshness: BClaude Code +9 - 0.518score
intake
Anubis44197
Interview the user, lock the writing brief, and define layout/front-matter/cover requirements before planning starts.
SkillBFreshness: BClaude Code +9 - 0.518score
run
Anubis44197
Single-command pipeline launcher for propose->design->create->polish->rewrite->export with hard gates.
SkillBFreshness: BClaude Code +9 - 0.518score
design
Anubis44197
Route ambiguous design requests to propose, design-big, design-small, or integrated sequence.
SkillBFreshness: BClaude Code +9 - 0.518score
design-big
Anubis44197
Produce approved full-book architecture: book plan, chapter plan, layout plan, state ledgers, and config scaffold.
SkillBFreshness: BClaude Code +9 - 0.513score
plot-hook
Anubis44197
Generate and refine plot architecture for arcs, chapter beats, dramatic turns, and resolution pacing.
SkillBFreshness: BClaude Code +9 - 0.513score
character
Anubis44197
Generate and refine character architecture documents for identity, relationships, and dialogue signatures.
SkillBFreshness: BClaude Code +9 - 0.513score
propose
Anubis44197
Generate three differentiated book proposals from genre, concept, audience, and publication-format constraints.
SkillBFreshness: BClaude Code +9 - 0.513score
bootstrap
Anubis44197
Generate and refine the bootstrap/source-of-truth document for world rules and concept constraints.
SkillBFreshness: BClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Jul 30, 2026 (3 months ago)
- Last release
- Aug 16, 2026 (2 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- Apache-2.0
- Repository
- github.com/Anubis44197/kit_hub
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from Anubis44197
- 0.523score
polish
Anubis44197
Polish chapters through book-structure diagnostics, editorial correction, TDK checks, and layout review.
SkillBFreshness: BClaude Code +9 - 0.522score
create
Anubis44197
Create book chapters sequentially using mapped design docs, guard rails, continuity tracking, and automated quality gates.
SkillBFreshness: BClaude Code +9 - 0.519score
export-word
Anubis44197
Export a complete validated book package to DOCX using approval, front matter, cover brief, and book-mode style gates.
SkillBFreshness: BClaude Code +9 - 0.518score
design-small
Anubis44197
Produce detailed chapter-range design using big-design outputs as prerequisites.
SkillBFreshness: BClaude Code +9 - 0.518score
intake
Anubis44197
Interview the user, lock the writing brief, and define layout/front-matter/cover requirements before planning starts.
SkillBFreshness: BClaude Code +9 - 0.518score
run
Anubis44197
Single-command pipeline launcher for propose->design->create->polish->rewrite->export with hard gates.
SkillBFreshness: BClaude Code +9
Cite this page
rewrite skill. Agent Mart. https://agentmart.dev/skills/anubis44197-rewrite (last updated 2026-10-24)