Agent skill

Hyperspacedb Cognitive

by YARlabs in YARlabs/hyperspace-db

Cognitive AI tools for HyperspaceDB: Chain-of-Thought stability analysis, Koopman momentum prediction, trust scoring, and Lyapunov convergence for agent reasoning.

MITAuto-check passedAgent Workflows

Install Hyperspacedb Cognitive

skills CLI
$ npx skills add YARlabs/hyperspace-db --skill hyperspacedb-cognitive -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install YARlabs/hyperspace-db hyperspacedb-cognitive --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/YARlabs/hyperspace-db.git skills-src && mkdir -p .claude/skills && cp -r skills-src/integrations/hyperspacedb-skills/skills/hyperspacedb-cognitive .claude/skills/hyperspacedb-cognitive && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
hyperspacedb-cognitive
GitHub stars
162
Token cost
~1.6k tokens
SKILL.md length
373 words
Files
1
Skills in repo
5
Repo updated
First seen
Licence
MIT

At a glance

Cognitive AI tools for HyperspaceDB: Chain-of-Thought stability analysis, Koopman momentum prediction, trust scoring, and Lyapunov convergence for agent reasoning.

  • Works in 4 steps: Lyapunov Thought Stability Analysis → Koopman Momentum Prediction → Trust Score → …
  • Working with AI agent memory
  • SKILL.md covers Core Concept: Thought…, 1. Lyapunov Thought Stability…, 2. Koopman Momentum Prediction and 3. Trust Score, plus 4 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Hyperspacedb Cognitive is an agent skill from YARlabs/hyperspace-db. Cognitive AI tools for HyperspaceDB: Chain-of-Thought stability analysis, Koopman momentum prediction, trust scoring, and Lyapunov convergence for agent reasoning. Use this skill when working with AI agent memory, reasoning stability, hallucination detection, thought trajectory forecasting, or Koopman operator theory. Trigger on: "thought stability", "chain of thought", "CoT", "hallucination detection", "reasoning loop", "Lyapunov", "momentum", "trust score", "agent memory", "attractor".

Its SKILL.md is about 1.6k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Agent Workflows, covering Agent memory and Forecasting and time series. The repository describes itself as: [H] HyperspaceDB is a high-performance, vector database. It features 1-bit quantization, async replication, and native support for hierarchical datasets (Lorentz, Poincaré ball &… The licence is MIT.

When your agent uses it

  • Working with AI agent memory
  • Reasoning stability
  • Hallucination detection
  • Thought trajectory forecasting

Example prompts

  • “thought stability”
  • “chain of thought”
  • “hallucination detection”
  • “/hyperspacedb-cognitive”

Workflow steps

4 steps, taken from the step headings in SKILL.md.

  1. Lyapunov Thought Stability Analysis
  2. Koopman Momentum Prediction
  3. Trust Score
  4. Gromov Delta / Geometry Analysis

What it can do on your machine

Read from SKILL.md and the folder at commit a43ccf7. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md (its code samples are typescript).

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Hyperspacedb Cognitive loads about 1.6k tokens when it runs. Until then it costs about 129 tokens; SKILL.md has 373 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~129
When it runs · the whole SKILL.md, loaded when a task matches
~1.6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from YARlabs/hyperspace-db at commit a43ccf7, republished under its MIT licence (© YARlabs). 373 words, ~1,570 tokens.

Download SKILL.mdSave it as .claude/skills/hyperspacedb-cognitive/SKILL.md (or your agent's skills folder).
name
hyperspacedb-cognitive
description
Cognitive AI tools for HyperspaceDB: Chain-of-Thought stability analysis, Koopman momentum prediction, trust scoring, and Lyapunov convergence for agent reasoning. Use this skill when working with AI agent memory, reasoning stability, hallucination detection, thought trajectory forecasting, or Koopman operator theory. Trigger on: "thought stability", "chain of thought", "CoT", "hallucination detection", "reasoning loop", "Lyapunov", "momentum", "trust score", "agent memory", "attractor".

HyperspaceDB Cognitive AI Tools

HyperspaceDB provides first-class cognitive primitives for AI agents. These tools are implemented as client-side computations in the SDK: they fetch stored vectors via getPoints() and then apply mathematical analysis locally. This means they work against any version of the HyperspaceDB server.

Supported geometry for cognitive tools:

  • lorentz, poincare — full Lorentz/hyperbolic math
  • hybrid — applies Lorentz math to first 33 dims, Euclidean to the rest
  • cosine, l2 — Euclidean approximations

Core Concept: Thought Trajectories

A thought trajectory is a sequence of vector IDs representing the progression of an agent's reasoning (e.g., each step of a Chain of Thought stored as a vector).

[id_1: "observe problem"] → [id_2: "form hypothesis"] → [id_3: "test hypothesis"] → ...

By storing CoT steps in HyperspaceDB, you can then apply mathematical analysis to detect hallucination, measure convergence, and predict future reasoning direction.


1. Lyapunov Thought Stability Analysis

Determines whether a reasoning trajectory is converging (stable attractor) or diverging (hallucination / reasoning loop).

Implementation: analyzeThoughtStability fetches the vectors for the given IDs via getPoints(), then computes Lyapunov exponent client-side in the SDK.

typescript
// trajectoryIds: ordered IDs of reasoning steps stored in the collection
const stability = await client.analyzeThoughtStability(
  trajectoryIds,      // number[] — ordered IDs of reasoning steps
  1.0,                // curvature parameter (1.0 for Lorentz/hybrid space)
  "reasoning_memory"
);
// returns: { lyapunov_exponent: number, is_stable: boolean, attractor_id?: number }

Interpretation:

lyapunov_exponentMeaning
< 0Stable — converging on a logical conclusion
≈ 0Neutral — bounded but not converging
> 0Unstable — potential hallucination or infinite loop

Use case: After each N steps of a reasoning chain, check stability. If unstable, trigger self-correction or inject a grounding prompt.


Show full SKILL.md (166 more words)Show less

2. Koopman Momentum Prediction

Extrapolates future reasoning direction using Koopman operator theory. Given a trajectory, predicts where the agent's thought process will move next.

Implementation: Fetches the last two trajectory vectors via getPoints(), calls Metric::extrapolate_momentum() client-side. For hybrid collections, Lorentz and Euclidean parts are extrapolated independently then recombined.

typescript
const forecast = await client.predictMomentum(
  trajectoryIds,      // number[] — past reasoning steps (min 2 IDs needed)
  1.0,                // steps ahead to predict
  "reasoning_memory",
  1.0                 // curvature
);
// returns: number[] — predicted next vector in the collection's geometry

Use case: Pre-fetch relevant context before the agent needs it, based on where its reasoning is heading — reducing latency in agentic loops.


3. Trust Score

Calculates a composite stability and coherence score (0.0 – 1.0) for a trajectory. Combines Lyapunov exponent with geometric consistency.

Implementation: Client-side computation on vectors fetched via getPoints().

typescript
const trust = await client.getTrustScore(
  trajectoryIds,      // number[]
  "reasoning_memory",
  1.0                 // curvature
);
// returns: number (0.0 – 1.0)

Use case: Gate critical decisions on trust score — only take action when reasoning confidence exceeds a threshold (e.g., trust > 0.75).


4. Gromov Delta / Geometry Analysis

Analyzes a set of raw vectors to determine the optimal database geometry. Ports the Gromov 4-point condition test.

typescript
import { CognitiveMathExport } from 'hyperspace-sdk-ts';

const { delta, recommendation } = CognitiveMathExport.analyzeDeltaHyperbolicity(
  vectorSamples,  // number[][]
  100             // numSamples
);
// recommendation: "lorentz" | "poincare" | "cosine" | "l2"

Pattern: Agentic Memory with Cognitive Feedback

typescript
// 1. Store reasoning steps as vectors
const stepIds: number[] = [];
for (const step of chainOfThought) {
  const id = await client.insertText(step.text, { step: step.index }, "agent_cot");
  stepIds.push(id);
}

// 2. Check stability after every 5 steps
if (stepIds.length % 5 === 0) {
  const { is_stable, lyapunov_exponent } = await client.analyzeThoughtStability(
    stepIds.slice(-10), 1.0, "agent_cot"
  );
  if (!is_stable) {
    // Inject grounding: retrieve most stable past conclusion
    const trustScores = await client.getTrustScore(stepIds.slice(-10), "agent_cot");
    console.warn(`Reasoning diverging (λ=${lyapunov_exponent}). Trust: ${trustScores.score}`);
  }
}

// 3. Predict next topic cluster
const forecast = await client.predictMomentum(stepIds, 1.0, "agent_cot");
// Pre-fetch related documents for the predicted direction
const upcoming = await client.search(forecast.predicted_vector, 5, "knowledge_base");

Math Utilities (Client-side)

typescript
import { CognitiveMathExport as CognitiveMath, HyperbolicMath } from 'hyperspace-sdk-ts';

// Lorentz inner product
const inner = CognitiveMath.lorentzInner(v1, v2);

// Geodesic distance in Lorentz space
const dist = CognitiveMath.lorentzDist(v1, v2);

// Project to Lorentz hyperboloid
const projected = HyperbolicMath.toLorentz(euclideanVec);

See Also

© YARlabs, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in integrations/hyperspacedb-skills/skills/hyperspacedb-cognitive of YARlabs/hyperspace-db.

Open the folder on GitHubat commit a43ccf7

Compare with similar skills

Hyperspacedb Cognitive next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Hyperspacedb Cognitive compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Hyperspacedb Cognitive this skillYARlabs/hyperspace-db162—~1.6kAutomated safety check: PassMIT
Insightsrohitg00/pro-workflow2.9k—~719Automated safety check: PassNone
Learning LoopLeoYeAI/openclaw-master-skills2.2k—~4.4kAutomated safety check: PassMIT
Tiktok Shop Inventorynexscope-ai/eCommerce-Skills1.1k—~444Automated safety check: PassMIT
TimesFM Forecastinggoogle-research/timesfm34k—~4.7kAutomated safety check: PassApache-2.0
Custom Mode Creator for claude-memthedotmack/claude-mem99k—~2.4kAutomated safety check: PassApache-2.0

Similar skills

  • Insights

    rohitg00/pro-workflow

    Show session analytics, learning patterns, correction trends, heatmaps, and productivity metrics.

    2.9k GitHub stars~719 tokensUpdated 12 days ago
    Agent WorkflowsAuto-check passed
  • Learning Loop

    LeoYeAI/openclaw-master-skills

    Structured self-improvement system for AI agents with confidence decay, cross-agent sharing, and anomaly detection.

    2.2k GitHub stars~4.4k tokensUpdated 2 mo ago
    Data & AnalyticsAuto-check passed
  • Tiktok Shop Inventory

    nexscope-ai/eCommerce-Skills

    Inventory management for TikTok Shop — demand forecasting, viral stock planning, FBT optimization

    1.1k GitHub stars~444 tokensUpdated 1 mo ago
    Data & AnalyticsAuto-check passed
  • TimesFM Forecasting

    google-research/timesfm

    Forecasts any univariate time series zero-shot with Google's TimesFM model, returning point forecasts and calibrated prediction intervals without training.

    34k GitHub stars~4.7k tokensUpdated 11 days ago
    Data & AnalyticsAuto-check passed
  • Walks you through creating, installing and verifying a custom claude-mem mode, including note types, tags and optional Telegram alerts for chosen memories.

    99k GitHub stars~2.4k tokensUpdated 2 days ago
    Agent WorkflowsAuto-check passed
  • Official

    Starts three parallel reviewer subagents over the current conversation transcript, then turns their findings into concrete edits to existing skills.

    11k GitHub starsUsed in 5 repos~1.2k tokens
    Agent WorkflowsAuto-check passed

More from YARlabs/hyperspace-db

  • Hyperspacedb Core

    YARlabs/hyperspace-db

    Core operations for HyperspaceDB — a multi-geometry vector database.

    162 GitHub stars~1.5k tokensUpdated yesterday
    Auto-check passed
  • Hyperspacedb Graph

    YARlabs/hyperspace-db

    Graph traversal, Lorentz hierarchy, and concept relationship operations for HyperspaceDB.

    162 GitHub stars~1.4k tokensUpdated yesterday
    Auto-check passed
  • Hyperspacedb MCP

    YARlabs/hyperspace-db

    Model Context Protocol (MCP) server for HyperspaceDB. An agent skill from YARlabs/hyperspace-db.

    162 GitHub stars~1.6k tokensUpdated yesterday
    Auto-check passed
  • Hyperspacedb Memory

    YARlabs/hyperspace-db

    Zero-overhead cognitive long-term and episodic memory for AI agents (Mem0 drop-in and mcp-hyperspace-memory).

    162 GitHub stars~1.2k tokensUpdated yesterday
    Auto-check passed

Questions about Hyperspacedb Cognitive

What does Hyperspacedb Cognitive do?

Cognitive AI tools for HyperspaceDB: Chain-of-Thought stability analysis, Koopman momentum prediction, trust scoring, and Lyapunov convergence for agent reasoning. Hyperspacedb Cognitive is an agent skill from YARlabs/hyperspace-db. Cognitive AI tools for HyperspaceDB: Chain-of-Thought stability analysis, Koopman momentum prediction, trust scoring, and Lyapunov convergence for agent reasoning.

When should I use Hyperspacedb Cognitive?

Hyperspacedb Cognitive fits situations like: working with AI agent memory; reasoning stability; hallucination detection; thought trajectory forecasting.

How do I install Hyperspacedb Cognitive in Claude Code?

Run `npx skills add YARlabs/hyperspace-db --skill hyperspacedb-cognitive -a claude-code`. Or copy the skill folder (integrations/hyperspacedb-skills/skills/hyperspacedb-cognitive in YARlabs/hyperspace-db) into .claude/skills/hyperspacedb-cognitive in your project. Claude Code loads it when a task matches its description.

How do I install Hyperspacedb Cognitive in Codex?

Run `npx skills add YARlabs/hyperspace-db --skill hyperspacedb-cognitive -a codex`. Or copy the skill folder (integrations/hyperspacedb-skills/skills/hyperspacedb-cognitive in YARlabs/hyperspace-db) into .agents/skills/hyperspacedb-cognitive in your project. Codex loads it when a task matches its description.

Can I use Hyperspacedb Cognitive in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add YARlabs/hyperspace-db --skill hyperspacedb-cognitive -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/hyperspacedb-cognitive, .gemini/skills/hyperspacedb-cognitive, .github/skills/hyperspacedb-cognitive and .opencode/skills/hyperspacedb-cognitive in your project.

What does Hyperspacedb Cognitive need to run?

SKILL.md names no scripts, command-line tools or credentials: Hyperspacedb Cognitive is instructions for the agent only.

Does Hyperspacedb Cognitive access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Hyperspacedb Cognitive safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Hyperspacedb Cognitive use?

Hyperspacedb Cognitive is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Hyperspacedb Cognitive use?

About 1.6k tokens (SKILL.md is roughly 6.3k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Hyperspacedb Cognitive?

Skills that share tags, products or a category with Hyperspacedb Cognitive: Insights (rohitg00/pro-workflow, 2.9k stars), Learning Loop (LeoYeAI/openclaw-master-skills, 2.2k stars), Tiktok Shop Inventory (nexscope-ai/eCommerce-Skills, 1.1k stars) and TimesFM Forecasting (google-research/timesfm, 34k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Hyperspacedb Cognitive?

YARlabs (a GitHub user) maintains it in YARlabs/hyperspace-db, which has 162 GitHub stars. The repository holds 5 skills in this directory. The repository was last updated on October 9, 2026.

Source: YARlabs/hyperspace-db on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.