agent-context-isolation

AI Community

Repo stars 3,783
Author repo Continuous-Claude-v3

Interpretation is structured for decision-making; original keeps the upstream SKILL.md unchanged.

Decide Fit First

Core job: Agent Context Isolation Prevent agent output from polluting the main context window. Task(subagenttype="...", runin_background=t…
Best fit: Use it when the task has reusable inputs, steps, and validation criteria rather than a one-off answer.
Avoid forcing it: If the source lacks commands, platform support, or external-service evidence, keep those fields unknown instead of guessing.

Design Intent

Structure: The skill is organized around “Rules”, “1. Use Background Agents with File-Based Coordination”, “2. Never Use TaskOutput to Retrieve Results”, “3. Monitor Agent Progress via System Reminders”, showing how the author expects the agent to judge fit, collect context, and produce verifiable output.
Trigger evidence: Prioritize the author’s wording around when to use it, what context to collect, and what output shape to produce.
Evidence boundary: Author text states facts, repository files prove commands and paths, and Fluxly only adds fit, limits, and usage judgment.

How To Use It

Inputs: Provide target material, scope, expected result, forbidden changes, and validation method.
Invocation: Name agent-context-isolation directly; if the source includes slash commands, start with the command and then add task context.
Validation: Start small and check whether the result follows “Rules / 1. Use Background Agents with File-Based Coordination / 2. Never Use TaskOutput to Retrieve Results” before expanding.

Boundaries And Review

Dependencies: It usually needs no extra API key, so start with a small validation task.
Permissions: Declared permissions include read / write; ask the agent to state file, command, and rollback boundaries before acting.
Quality bar: A useful result names the deliverable, evidence, and next action. Generic prose means the task needs tighter context.

Agent Context Isolation

Prevent agent output from polluting the main context window.

Rules

1. Use Background Agents with File-Based Coordination

# RIGHT - background agent writes to file, main reads file
Task(subagent_type="...", run_in_background=true, prompt="... Output to: /path/to/file.md")

# WRONG - foreground agent dumps full transcript into main context
Task(subagent_type="...", run_in_background=false)

Background agents with run_in_background=true isolate their context. Have them write results to files in .claude/cache/agents/<agent-type>/.

2. Never Use TaskOutput to Retrieve Results

# WRONG - dumps entire transcript (70k+ tokens) into context
TaskOutput(task_id="<id>")
TaskOutput(task_id="<id>", block=true)

# RIGHT - check expected output files
Bash("ls -la .claude/cache/agents/<agent-type>/")
Bash("bun test")  # verify with tests

TaskOutput returns the full agent transcript. Always use file-based coordination instead.

3. Monitor Agent Progress via System Reminders

# System reminders come automatically:
# "Agent a42a16e progress: 6 new tools used, 88914 new tokens"

# To detect completion:
# - Watch for progress reminders to stop arriving
# - Poll for expected output files: find .claude/cache/agents -name "*.md" -mmin -5
# - Check task output file size growth: wc -c /tmp/claude/.../tasks/<id>.output

Stuck agent detection:

Progress reminders stop arriving
Task output file size stops growing
Expected output file not created after reasonable time

4. Verify with Tests, Not Output

After agent work:

Run the test suite directly: bun test
Report pass/fail counts
Only investigate failures if tests fail

5. File-Based Agent Pipeline Pattern

Research agent → .claude/cache/agents/oracle/output.md
                          ↓
Plan agent → .claude/cache/agents/plan-agent/output.md (reads research)
                          ↓
Validate agent → .claude/cache/agents/validate-agent/output.md (reads plan)
                          ↓
Implement agent → src/module.ts (reads validated plan)

Each agent reads the previous agent's file output, not TaskOutput.

Why This Matters

Agent context isolation preserves the main conversation's context budget. Reading agent outputs via TaskOutput floods context, causing:

Mid-conversation compaction
Lost context about user's original request
Repeated explanations needed

Source

Session where TaskOutput flooded 70k+ tokens into main context
Session 2026-01-01: Successfully used background agents with file-based coordination for SDK Phase 3

Fluxly profile Author and license come from source; runtime, permissions, and network are Fluxly detections or estimates

Fluxly category: AI
Author-declared agents: No explicit declaration found; this is not inferred or tested compatibility
Static check: 88 / 100 · heuristic scan, not runtime safety proof
Author / version / license: @parcadei · no license declared
Fluxly token estimate: Lean
Fluxly setup estimate: Plug-and-play
External API key: No requirement detected
Detected OS requirements: macOS · Linux · Windows
Runtime requirements: Unspecified
Detected file/system behavior: Read-only
Write / modify
Detected network behavior: Local-only
Install commands: None (reference only)

Profile is derived at build time from SKILL.md and install vectors. Subject to drift from author intent.

Heads up: 未限定 allowed-tools，默认拥有全部工具权限。

Output preview agent-context-isolation.preview

# 2. Never Use TaskOutput to Retrieve Results

# WRONG - dumps entire transcript (70k+ tokens) into context
TaskOutput(task_id="<id>")
TaskOutput(task_id="<id>", block=true)

# RIGHT - check expected output files
Bash("ls -la .claude/cache/agents/<agent-type>/")
Bash("bun test")  # verify with tests

SKILL.md SKILL.md

# Agent Context Isolation

Prevent agent output from polluting the main context window.

## Rules

### 1. Use Background Agents with File-Based Coordination
```
# RIGHT - background agent writes to file, main reads file
Task(subagent_type="...", run_in_background=true, prompt="... Output to: /path/to/file.md")

# WRONG - foreground agent dumps full transcript into main context
Task(subagent_type="...", run_in_background=false)
```

Background agents with `run_in_background=true` isolate their context. Have them write results to files in `.claude/cache/agents/<agent-type>/`.

### 2. Never Use TaskOutput to Retrieve Results
```
# WRONG - dumps entire transcript (70k+ tokens) into context
TaskOutput(task_id="<id>")
TaskOutput(task_id="<id>", block=true)

# RIGHT - check expected output files
Bash("ls -la .claude/cache/agents/<agent-type>/")
Bash("bun test")  # verify with tests
```

TaskOutput returns the full agent transcript. Always use file-based coordination instead.

### 3. Monitor Agent Progress via System Reminders
```
# System reminders come automatically:
# "Agent a42a16e progress: 6 new tools used, 88914 new tokens"

# To detect completion:
# - Watch for progress reminders to stop arriving
# - Poll for expected output files: find .claude/cache/agents -name "*.md" -mmin -5
# - Check task output file size growth: wc -c /tmp/claude/.../tasks/<id>.output
```

**Stuck agent detection:**
1. Progress reminders stop arriving
2. Task output file size stops growing
3. Expected output file not created after reasonable time

### 4. Verify with Tests, Not Output
After agent work:
1. Run the test suite directly: `bun test`
2. Report pass/fail counts
3. Only investigate failures if tests fail

### 5. File-Based Agent Pipeline Pattern
```

…

Evidence boundary and execution chain

Author text anchors workflow facts; Fluxly only indexes current sections, terms, files, and commands.

sections -> Rules → 1. Use Background Agents with File-Based Coordination → 2. Never Use TaskOutput to Retrieve Results → 3. Monitor Agent Progress via System Reminders → 4. Verify with Tests, Not Output → 5. File-Based Agent Pipeline Pattern
terms    -> Stuck agent detection · Prevent agent output from polluting the main context window. · Background agents with runinbackground=true isolate their context. · TaskOutput returns the full agent transcript. · Stuck agent detection: 1. · Each agent reads the previous agent's file output, not TaskOutput. · Agent context isolation preserves the main conversation's context budget.
files/cmd -> runinbackground=true · .claude/cache/agents/<agent-type>/ · bun test · path/to/file.md · .claude/cache/agents · tmp/claude/.../tasks · pass/fail · .claude/cache/agents/oracle/output.md
body sha256 -> 0a563b12cba7