/Catalogue/Prompt/parcadei/parcadei-continuous-claude-v3-agent-context-isolation

Origin: github

agent-context-isolation

Agent Context Isolation

by parcadei · updated 8mo ago · imported from GitHub

Installs0+0/7d
Security score100/100
Retention 14d0%
GitHub stars3.9K

Skill logic

Execution graph
User message
Prompt rewrites behaviour
Response

SKILL.md

View on GitHub ↗

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

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

Discussion

No comments yet — start the thread.

Sign in to join the discussion.

/More from parcadei/Continuous-Claude-v3

parcadei· 8mo agoCommunity
dead-code

Prompts · Python · v0.1.0

Find unused functions and dead code in the codebase

#agents#claude-code#claude-code-cli

0 3.9K
parcadei· 8mo agoCommunity
agent-orchestration

Prompts · Python · v0.1.0

Agent Orchestration Rules

#agents#claude-code#claude-code-cli

0 3.9K
parcadei· 8mo agoCommunity
agentic-workflow

Prompts · Python · v0.1.0

Agentic Workflow Pattern

#agents#claude-code#claude-code-cli

0 3.9K
parcadei· 8mo agoCommunity
agentica-claude-proxy

Prompts · Python · v0.1.0

Guide for integrating Agentica SDK with Claude Code CLI proxy

#agents#claude-code#claude-code-cli

0 3.9K