How does Context Recovery find lost information?

It uses the aspy proxy's recall tools to perform semantic and keyword searches through your interaction history, including prompts, responses, and internal thinking blocks.

Does it work across different coding sessions?

Yes, it is specifically built to retrieve information from both the current compacted session and previous development sessions stored in the aspy history.

What is session compaction in Claude Code?

Session compaction occurs when a conversation becomes too long for the AI's context window, causing older parts of the chat to be summarized or removed to save token space.

What makes its synthesis better than a raw search?

Instead of dumping raw logs, it synthesizes findings into actionable summaries, highlighting specific file paths, line numbers, and decided-upon implementation patterns.

Can I search for vague topics using this skill?

Yes, the skill is designed to handle fuzzy queries. Because it uses semantic search, it can find conceptually related content even if you don't remember the exact keywords used.

Context Recovery

Name: Context Recovery
Author: omgpointless

byomgpointless

•

Productivity & Workflow

Recovers and synthesizes lost session context using semantic search across previous conversations, thinking blocks, and assistant responses.

About

The Context Recovery skill enables Claude to regain critical information lost during session compaction or across separate coding sessions. By utilizing the aspy proxy's recall capabilities, it performs semantic and keyword-based searches through past user prompts, Claude's internal reasoning (thinking blocks), and historical code implementations. This ensures seamless continuity in development by helping the AI remember previous decisions, specific file locations, and unfinished tasks without requiring the user to manually re-explain their progress.

Key Features

1 GitHub stars
Fuzzy query handling for vague references like 'that golf thing'
Semantic and keyword-based search across entire session history
Deep search within thinking blocks to retrieve technical rationale
Targeted recall tools for prompts, responses, or internal reasoning
Automatic synthesis of findings into concise summaries with file paths

Use Cases

Resuming a complex feature implementation after session compaction has cleared the context
Recalling specific architectural decisions or bug fixes made in a previous day's session
Identifying 'next steps' or unfinished work mentioned earlier in a long-running project

About

Key Features

1 GitHub stars
Fuzzy query handling for vague references like 'that golf thing'
Semantic and keyword-based search across entire session history
Deep search within thinking blocks to retrieve technical rationale
Targeted recall tools for prompts, responses, or internal reasoning
Automatic synthesis of findings into concise summaries with file paths

Use Cases

Resuming a complex feature implementation after session compaction has cleared the context
Recalling specific architectural decisions or bug fixes made in a previous day's session
Identifying 'next steps' or unfinished work mentioned earlier in a long-running project