Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/vima-tech/innate/assetsnpx skills add vima-tech/Innate --skill assetsgit clone --depth 1 https://github.com/vima-tech/InnateWhat 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 | $0.00156 | $0.04112 |
| Opus 5 | $0.00078 | $0.02056 |
| Sonnet 5 | $0.00031 | $0.00822 |
| Haiku 4.5 | $0.00016 | $0.00411 |
Grade A, and why
innate-memory 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 2d 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.
Copies of this mod
1 near-identical copy found in the catalogue:
- innate-memory — 100% identical, 0 lines differ
How it starts
The opening of the file, as written. The whole thing — 368 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Layer Role
Innate is auxiliary — it must never block the main task. All innate operations are best-effort. On any tool error: retry once with a corrected call, then abandon and continue.
When to Recall
Recall at the start of a task only if the task is:
- In a domain with known recurring patterns (e.g., a specific codebase, framework, or protocol)
- About debugging or fixing something that may have been encountered before
- Architectural/design — prior decisions matter
Skip recall for: quick one-liners, questions with clear answers, tasks the user scoped as entirely new territory with no prior context.
Query Formulation
Use canonical intent, not literal user words. Keep queries consistent across sessions — the same class of problem should always use the same canonical form, because Innate accumulates context statistics per query pattern.
| User says | Good query | Bad query |
|---|---|---|
| "why does this crash on startup?" | startup crash sqlite init |
why does this crash on startup |
| "refactor the auth flow" | auth flow session token handling |
refactor auth flow |
| "add rate limiting to the API" | rate limiting middleware pattern |
add rate limiting |
| "debug the flaky test" | flaky test race condition timing |
debug flaky test |
Call: innate_recall(query=<canonical_intent>, budget=4000, source="mcp")
Inject recalled chunks into your working context. Mention to the user which (if any) relevant prior knowledge was found — one sentence is enough.
Auto-Recall Hook
When Innate is installed with hooks, a UserPromptSubmit hook may automatically inject an
<innate-recall> block into the conversation — it already ran a relevance-gated recall for the
prompt and carries a trace_id. When you see that block:
- Reuse its
trace_idfor the closinginnate_record— do not callinnate_recallagain for the same task (that would create a duplicate, dangling trace). - Treat its chunks exactly like ones you recalled yourself: apply what helps, then close with a per-chunk verdict (see below).
- If the block is absent and the task warrants it, recall manually as above.
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 2d ago First seen · 368 lines · 156 tokens per session scan A f747896331e2
innate-memory is a skill published in the GitHub repository vima-tech/Innate (2 stars, last pushed 25d ago), licensed MIT. It adds 156 tokens to every session and 4,112 once invoked, about $0.0008 per session on Opus 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 2026-08-31.
Other skills, from other repositories
kayba-stage-3-metrics
Define metrics from Kayba insights, implement them as Python measurement code, run against traces, and iterate until the metrics are clean and meaningful. Trigger when the user says "run stage 3", "define metrics", "build metrics", "compute baselines", or when invoked by the kayba-pipeline orchestrator. Requires…
kayba-stage-5-action-plan
Triage each insight into discard/code-fix/prompt-fix and produce a prioritized action plan with specific recommendations. Trigger when the user says "run stage 5", "make action plan", "triage skills", or when invoked by the kayba-pipeline orchestrator. Requires eval outputs from stages 1-4.
kayba-stage-4-rubric
Organize computed metrics into a tiered evaluation rubric with leading, lagging, and quality indicators. Trigger when the user says "run stage 4", "build rubric", "tier metrics", or when invoked by the kayba-pipeline orchestrator. Requires eval/baselinemetrics.json and eval/computebaselines.py to exist.
kayba-stage-6-hitl
Human-In-The-Loop gate that presents the action plan with full context, collects an informed approval/modification/rejection decision, and records the outcome. Trigger when the user says "run stage 6", "HITL review", "approve action plan", or when invoked by the kayba-pipeline orchestrator. Requires eval/actionplan.md…
kayba-pipeline
End-to-end agent evaluation and improvement pipeline. Takes a traces folder and optional HITL flag, then orchestrates sub-agents through 7 stages — each stage is its own skill invoked by a dedicated sub-agent. Trigger when the user says "run the pipeline", "kayba pipeline", "evaluate and fix", "full eval", "analyze…
kayba-stage-2-domain-context
Gather domain context about the repository and agent — system prompt, tool definitions, domain docs, and behavior patterns from traces. Trigger when the user says "run stage 2", "gather context", "domain context", or when invoked by the kayba-pipeline orchestrator.