inkos

InkOS is a Node.js/TypeScript CLI/agent tool that autonomously generates, audits, and revises novels chapter-by-chapter using a multi-agent pipeline (plan/compose/draft/audit/revise), with continuity tracking via “truth” files, schema-validated JSON state deltas, and rollback/versioning. It also includes an optional local web studio and an OpenClaw-compatible skill interface with structured (e.g., --json) command outputs.

Evaluated Mar 30, 2026 (0d ago)
Repo ↗ Ai Ml cli ai-agent multi-agent novel-writing typescript continuity audit revision openclaw state-management sqlite
⚙ Agent Friendliness
68
/ 100
Can an agent use this?
🔒 Security
51
/ 100
Is it safe for agents?
⚡ Reliability
40
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
0
Documentation
78
Error Messages
--
Auth Simplicity
85
Rate Limits
20

🔒 Security

TLS Enforcement
60
Auth Strength
70
Scope Granularity
15
Dep. Hygiene
35
Secret Handling
65

CLI uses API keys via environment variables and stores global config in ~/.inkos/.env. No evidence provided about redaction/logging of secrets in the README, but separate mention of JSON Lines logs and silence mode suggests some logging controls may exist. Webhook support mentions HMAC-SHA256 signing (good for integrity). Scope granularity is not applicable like OAuth scopes (single upstream API key), and rate limit behavior is not documented. TLS enforcement to upstream is not explicitly stated, though base URLs imply HTTPS in typical usage.

⚡ Reliability

Uptime/SLA
0
Version Stability
45
Breaking Changes
30
Error Recovery
85
AF Security Reliability

Best When

You want CLI-driven, continuity-aware novel generation with audit/revision loops, prefer local state management (truth files/JSON deltas), and will integrate it into agent workflows via atomic commands and --json outputs.

Avoid When

You need a standardized web API with documented error contracts and pagination semantics, or you cannot handle local filesystem state (project directories, logs, SQLite DB) and LLM credential management.

Use Cases

  • Generating chapters for long-form novels with continuity across chapters
  • Author-guided writing using control documents (author_intent.md, current_focus.md) and chapter intents
  • Automated quality gates: audit dimensions, iterative revision loops, and human approval workflows (e.g., review approve-all)
  • Style transfer: analyze a sample text to extract writing fingerprints and apply them to a book
  • Continuation of existing novels by importing chapters and reconstructing state
  • Fanfic creation with mode controls (canon/au/ooc/cp) and boundary management
  • Scheduling autonomous writing via a background guard process (inkos up) with notifications and webhooks
  • Using as a toolset for other agent frameworks via exec/atomic commands

Not For

  • Use as a general-purpose REST/GraphQL backend API for applications
  • Use when you cannot provide/secure access to LLM provider credentials (OpenAI-compatible API key required)
  • Use where you need strict guarantees of content originality or legal clearance (no stated IP/compliance workflow in provided text)
  • Use when you require public enterprise-grade reliability guarantees (SLA/status, etc. not provided)

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
No
SDK
No
Webhooks
Yes

Authentication

Methods: API key for LLM providers (OpenAI-compatible via provider/custom/base-url) Optional env-var routing for per-agent model credentials
OAuth: No Scopes: No

Authentication is primarily delegation to the upstream LLM provider using INKOS_LLM_API_KEY / api-key-env. The tool itself is local CLI; no user signup described. For notifications, webhooks mention HMAC-SHA256 signing, but no user-facing auth for the CLI is described.

Pricing

Free tier: No
Requires CC: No

No pricing for InkOS itself is provided in the supplied text; costs depend on the configured LLM provider usage.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Requires correct local project initialization and editing/feeding control documents (author_intent.md, current_focus.md) to get predictable outcomes.
  • Atomic commands support --json, but orchestration should account for multi-file state compilation (plan/compose produce runtime artifacts used by subsequent steps).
  • Long-running guard process (inkos up) may pause for human review on key issues; agents should handle that control-flow.
  • Provider compatibility: while it supports OpenAI-compatible endpoints, not all intermediaries may support streaming/SSE; tool claims fallback but agent logic should expect mode differences.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for inkos.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

6026
Packages Evaluated
20377
Need Evaluation
586
Need Re-evaluation
Community Powered