Summarize and compress the current conversation history into a structured context snapshot, then call compact_chat_history to save it. Read this skill only when the user explicitly asks to compact/summarize — system-triggered compaction injects the instructions directly without requiring a skill read.
64
75%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
View guide
High
Do not use without reviewing
Fix and improve this skill with Tessl
tessl review fix ./backend/super-magic/agents/skills/compact-chat-history/SKILL.mdCompress the current conversation into a structured summary, then call compact_chat_history immediately with the result.
The conversation history is too long and must be compressed. You must call the compact_chat_history tool immediately to complete the summary.
Your task is to create a thorough summary of the conversation so far, with special attention to the user's explicit requests and your prior actions. The summary must capture all important details, work results, and file locations to ensure continuity, since we follow an "everything-is-a-file" architecture.
Remember: subsequent work will restore context by reading files, so you must provide accurate file paths. For content already saved to files, note the location — do not repeat large blocks of text in the summary itself.
Your summary must include the following sections:
1. Task Goals and Approach
2. Key Files and Context Resources (most important part — be thorough and precise)
magic.project.js)3. Skills Needed to Resume This Task
4. Resolved Issues and Current State
5. Incomplete Tasks, Next Steps, and Continuity Confirmation
6. High-Value User Input
If any of the above sections overlap, merge them — no need to repeat.
1. Task Goals and Approach:
[Describe each specific request in detail]
- [Method 1]
- [Strategy 2]
- [...]
2. Key Files and Context Resources (most critical):
- [project outline path] - overall plan and structure - read when confirming global goals and scope
- [currently active file path] - current progress and key context - read first when resuming
- [user-specified reference path] - content user explicitly requested - read when working on that section
- [similar completed content path] - style/structure reference - read a suitable amount when consistency is needed
- [project config path] - project settings - read when config details are needed
- [history/backup path] - read when tracing back changes
- [info name] - not saved to file - how to re-acquire: [specific method]
Reading principles:
- Start with the most important items closest to the current task, then read others as needed
- For high-consistency tasks, read a suitable number of completed items as reference
- Avoid reading all files at once to prevent filling up context again
3. Skills Needed to Resume This Task:
- [High] [Skill name] - [purpose]
- [Medium] [Skill name] - [purpose]
- [Low] [Skill name] - [purpose]
4. Resolved Issues and Current State:
[Description of resolved issues and ongoing troubleshooting]
[Accurate description of current work state]
5. Incomplete Tasks, Next Steps, and Continuity Confirmation:
- Task 1
- Task 2
- [...]
[Optional next action plan]
[Current task and progress, or task completion notice]
6. High-Value User Input:
Verbatim quotes of user messages that are valuable — complete and unalteredsummary parameter of the compact_chat_history tool call to ensure complete delivery.41d7ef4
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.