Nm Conserve Compression Strategy

Analyze current context and recommend compression strategies for bloated or quota-heavy sessions

Security Scan

The skill's instructions are coherent with a context-compression purpose, but they reference reading/writing .claude files and spawning agents without declaring those config paths or required platform capabilities, which is an unexplained mismatch you should verify before installing.

Purpose & Capability

The skill's stated purpose (analyzing and recommending compression strategies) matches the tasks described (analyze /context, recommend strategies, estimate savings). However, the SKILL.md repeatedly reads/writes files under .claude (e.g., .claude/session-state.md, .claude/context-archive/) and invokes other platform capabilities (Skill(...) and Task tool) while the skill metadata declares no required config paths, credentials, or platform features. That discrepancy (file IO and inter-skill invocation not declared) is unexplained.

Instruction Scope

Instructions tell the agent to run platform commands (/context, /clear, /catchup), spawn agents (Skill(conserve:clear-context), Task tool), and read/write .claude/ files and archives. These operations go beyond a passive recommendation: they imply the agent will modify and persist session state. The SKILL.md does not limit or justify access to arbitrary .claude files, nor does it declare which exact paths it will read versus write.

Install Mechanism

No install spec and no code files are present; this is instruction-only so nothing will be downloaded or written at install time. That minimizes install-time risk.

Credentials

The skill requests no environment variables or external credentials, which is proportionate. However, it still performs file IO under a user-config directory; the lack of declared required config paths for those files is a mismatch the user should verify (i.e., the skill will read/write .claude/ even though no paths were declared).

Persistence & Privilege

always:false and no special privileges are requested. The skill instructs spawning agents and clearing context, which are normal platform operations but can be impactful (destructive /clear). There is no automatic permanent presence or config modification of other skills declared.

What to consider before installing

This is an instruction-only skill that mostly does what it says (analyze context and recommend compression strategies), but before installing or invoking it: 1) verify your platform supports the commands/tools it references (/context, /clear, /catchup, Skill(...) and Task tool); 2) confirm you are comfortable with the skill reading and writing files under .claude/ (it will save session-state and create context-archive files) — the skill metadata did not declare those config paths; 3) back up any important session data before allowing it to run /clear or similar destructive operations; 4) if you need stronger assurance, ask the publisher to explicitly declare the config paths and exact file IO behavior (read vs write), or provide a minimal test run mode that does not perform writes. If the author updates the metadata to declare .claude/* paths or documents sandboxing for file IO, the mismatch would be resolved and my confidence would increase.

Night Market Skill — ported from claude-night-market/conserve. For the full experience with agents, hooks, and commands, install the Claude Code plugin.

Compression Strategy

Analyze current context usage and recommend optimal compression strategies.

When To Use

• Context feels bloated or sluggish
• Before major task phase transitions (plan complete, starting implementation)
• Token quota burning faster than expected
• After large tool output accumulations

When NOT To Use

• Context-optimization skill already handling the scenario
• Simple queries with minimal context
• Freshly cleared context

Required TodoWrite Items

1. compression-strategy:analyze-context
2. compression-strategy:recommend-strategy
3. compression-strategy:estimate-savings

Step 1 – Analyze Context (analyze-context)

Run /context to check current usage. Then estimate:

1. Tool output accumulation: How much context is from tool results vs. conversation?
2. Stale content age: How many turns since critical decisions were made?
3. Active files: Which files are still relevant vs. historical?

Step 2 – Recommend Strategy (recommend-strategy)

Based on analysis, recommend one of:

Option A: /clear + /catchup

Best when:

• Task phase complete (planning done, implementation starting)
• Context > 60% full
• Most content is stale

Process:

1. Save critical state to .claude/session-state.md
2. Run /clear
3. Run /catchup to reload active files

Option B: Spawn Continuation Agent

Best when:

• Context > 80% full
• Work in progress, can't stop
• Delegatable tasks remain

Process:

1. Run Skill(conserve:clear-context) to spawn continuation agent
2. Agent receives fresh context with saved state

Option C: Archive + Summarize

Best when:

• Context 40-60% full
• Some stale content mixed with active
• Not ready for full clear

Process:

1. Archive old decisions/errors to .claude/context-archive/
2. Summarize completed work in memory
3. Continue with leaner context

Option D: Delegate to Subagent

Best when:

• Parallel work possible
• Independent subtasks exist
• Context pressure moderate

Process:

1. Identify delegatable tasks
2. Spawn specialized agents via Task tool
3. Main context stays lean

Step 3 – Estimate Savings (estimate-savings)

For the recommended strategy, estimate:

Strategy	Typical Savings	Risk
/clear + /catchup	70-90%	Low if state saved
Continuation agent	80-95%	Low, state preserved
Archive + summarize	20-40%	Very low
Delegate to subagent	30-50%	Low, parallel work

Context Archive Location

Preserved context is saved to:

.claude/context-archive/pre-compact-YYYYMMDD-HHMMSS-SESSIONID.md

This is automatically created by the pre_compact_preserve hook before any /compact operation.

Integration Points

• PreCompact hook: Automatically preserves context before compression
• Tool output summarizer: Warns when tool outputs accumulate
• Context warning hook: Three-tier alerts at 40%/50%/80%

Example Usage

/compression-strategy

Output:

Context Analysis:
- Current usage: 52%
- Tool output: ~15KB (3 tool results)
- Stale content: ~40% (decisions from 8+ turns ago)

Recommendation: Option C - Archive + Summarize
- Archive old decisions to context-archive
- Keep active files and recent decisions
- Estimated savings: 25-35%

Commands:
1. Read .claude/context-archive/ to see what's preserved
2. Summarize completed work
3. Continue with leaner context