forgetful-explore

forgetful-explore is a skill for Claude Code from ScottRBK/forgetful. It costs 65 tokens per session (813 once invoked), scanned A, original, MIT.

A workflow for exploring a Forgetful knowledge graph when a simple keyword search is not enough. It follows connections between memories, entities, and relationships to build a connected explanation.

In plain words
What is it for?
Use it to investigate what is known about an entity or topic, trace related decisions and memories, and choose shallow, medium, or deep exploration based on the question.
Why use it?
It helps answer questions that span projects or involve several related things. Tracking visited items prevents the investigation from repeatedly following the same branches.

Skill for Claude Code

Written for Claude Code: allowed-tools in frontmatter.

Install

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.

agentmods
npx agentmods add skills/scottrbk/forgetful/forgetful-explore
Any agent
npx skills add ScottRBK/forgetful --skill forgetful-explore
Clone the repo
git clone --depth 1 https://github.com/ScottRBK/forgetful

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for forgetful-explore

README.md
[![agentmods](https://agentmods.dev/badge/skills/scottrbk/forgetful/forgetful-explore.svg)](https://agentmods.dev/skills/scottrbk/forgetful/forgetful-explore)
Your own site
<a href="https://agentmods.dev/skills/scottrbk/forgetful/forgetful-explore"><img src="https://agentmods.dev/badge/skills/scottrbk/forgetful/forgetful-explore.svg" alt="Measured on agentmods" height="20"></a>
Per session 65 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 813 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What 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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00065 $0.00813
Opus 5 $0.00032 $0.00407
Sonnet 5 $0.00013 $0.00163
Haiku 4.5 $0.00006 $0.00081

Measured 6d ago against content hash 7093cf2b3355, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

forgetful-explore 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 6d 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.

skills/forgetful-explore/SKILL.md · 87 lines

How it starts

The opening of the file, as written. The whole thing — 87 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Exploring the knowledge graph

A flat query answers "what matches"; exploration answers "how it connects". The output is a synthesized picture that states the path taken — never a raw dump of hops.

Invoking operations

Operations are named by registry name (search_entities, get_entity_memories, ...). Invoke via whichever surface this agent has:

  • MCP: execute_forgetful_tool(tool_name="search_entities", arguments={...})
  • CLI: forgetful call search_entities --args '{"query": "..."}' --json

Get any operation's schema at runtime: how_to_use_forgetful_tool (MCP) or forgetful tools info <operation> (CLI) — schemas are deliberately not repeated here.

Choose a depth first

  • Shallow (phases 1–2): confirm or refresh something specific.
  • Medium (phases 1–4): understand a topic and its immediate neighbourhood.
  • Deep (all phases): cross-project investigation or "everything we know about X".

Track visited memory and entity IDs throughout — a revisited ID means that branch is done.

Phase 1 — Semantic entry

Broad query_memory on the topic (with query_context stating the investigation's intent). The goal is entry points, so favour breadth over precision here — raise k above its default of 3 (max 20) rather than settle for a narrow first pass.

Done when: a handful of strong entry memories are identified.

Phase 2 — Expand the strongest hits

get_memory on each entry point for full content and linked memory IDs; follow the links that bear on the question.

Done when: each entry point's local cluster is understood.

Phase 3 — Discover the entities

Collect entities referenced by the expanded memories, and search_entities for the topic's obvious actors (systems, people, components) that memories may not name directly.

Done when: the cast of relevant entities is listed.

Phase 4 — Walk the relationships

get_entity_relationships on each relevant entity; follow the relationship types that answer the question (depends_on for impact, part_of for structure, owns for responsibility).

Read the full file on GitHub · 87 lines

Changes

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.

  1. 6d ago First seen · 87 lines · 65 tokens per session scan A 7093cf2b3355

Subscribe to this mod's changes

forgetful-explore is a skill published in the GitHub repository ScottRBK/forgetful (298 stars, last pushed 4d ago), licensed MIT. It adds 65 tokens to every session and 813 once invoked, about $0.0003 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 2026-08-30.