BA novel-forge
Long-form novel workflow for creating, continuing, resuming, and repairing serialized fiction with externalized project state, role-to-model mapping, worldbuilding, character sheets, full outlines, 10-chapter batch outlines, style sampling, chapter drafting, consistency review, memory tracking, and spawned multi-session collaboration. Use when the user asks to start a novel project, continue or resume a draft, recover from truncation, assign models to roles, generate canon or chapters, review for consistency, or maintain a long-running fiction project across many chapters. Supports single-agent or multi-agent execution, with multi-agent as the default; when multi-agent is selected, first surface the available model inventory and the novel-writing role list, then ask the user for an explicit role→model mapping before any canon work. Once the user has provided the mapping, persist it in project state and drive stage work with `sessions_spawn` using the mapped roles rather than treating the mapping as passive metadata. The main session may only create the project shell and route work; it must not author canon files.
As a process A 88/100 · Runs to the end — weak spots: running it twice, progress reporting
How to improve
- Shorten the description to 1024 characters.
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 0
✓ No critical or high findings
Files scanned: 14. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- error
description-longdescription is 1130 chars, limit 1024
Process rating: all ten parameters 88/100
- 0Progress reporting. Says nothing while it works
- 30Running it twice. 5 mutating operations with no state check
- 60Tools and files. Uses tools (bash) that frontmatter does not declare
- 100Steps. 159 steps
- 100Result and completion. Output format and completion criterion are stated
- 100When it triggers. States when to use and when not to
- 100Inputs and preconditions. Inputs and preconditions are listed
- 100Failures and branches. 19 branches, has a failure section
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 3055 tokens
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
- low 12 top-level sections: this looks like several domains in one skill
- high The skill tells the model to perform an irreversible action with no human approval
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- +3Description length 1130: 120–800 characters recommended
- +4No input/output examples
- +1No license
- +2Single-language instructions
- +4Structure: 24 headings
- +3Step-by-step instructions: 159 items
- +3Output format is stated explicitly
- +4Reference files are cited in the instructions (7 of 8)
- +3All 4 scripts are documented
Quality base 70; lint remarks subtract, signals add up to 100. Result: 62.