
Cc Skill Continuous Learning
- 545 installs
- 44k repo stars
- Updated July 27, 2026
- sickn33/antigravity-awesome-skills
cc-skill-continuous-learning is a Claude Code Stop-hook skill that extracts reusable debugging and project patterns from finished sessions into ~/.claude/skills/learned/ for developers who want agents to remember proven
About
cc-skill-continuous-learning is a Claude Code agent skill that runs a bash evaluator on the Stop hook when a session ends, scanning transcripts for five pattern types—error_resolution, user_corrections, workarounds, debugging_techniques, and project_specific—while ignoring simple_typos, one_time_fixes, and external_api_issues. Configurable JSON sets min_session_length to 10 messages, extraction_threshold to medium, and auto_approve to false so developers review before skills land in ~/.claude/skills/learned/. The design favors Stop over UserPromptSubmit to avoid per-message latency. Reach for cc-skill-continuous-learning when Claude Code sessions repeatedly solve the same stack-specific bugs and you want those resolutions promoted into reusable SKILL.md files without manual copy-paste.
- Stop hook runs once per session end—lighter than UserPromptSubmit on every message
- Detects error_resolution, user_corrections, workarounds, debugging_techniques, and project_specific patterns
- Ignores simple_typos, one_time_fixes, and external_api_issues to reduce noise
- Configurable min_session_length default 10 with medium extraction_threshold and auto_approve false
- Writes extracted skills to ~/.claude/skills/learned/ with evaluate-session.sh orchestration
Cc Skill Continuous Learning by the numbers
- 545 all-time installs (skills.sh)
- Ranked #94 of 826 Skill Development skills by installs in the Skillselion catalog
- Security screen: HIGH risk (skills.sh audit)
- Data as of Jul 28, 2026 (Skillselion catalog sync)
npx skills add https://github.com/sickn33/antigravity-awesome-skills --skill cc-skill-continuous-learningAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Installs | 545 |
|---|---|
| repo stars | ★ 44k |
| Security audit | 2 / 3 scanners passed |
| Last updated | July 27, 2026 |
| Repository | sickn33/antigravity-awesome-skills ↗ |
How do you persist debugging patterns from Claude Code sessions?
Capture reusable debugging and project patterns from finished Claude Code sessions into ~/.claude/skills/learned/ via a Stop hook evaluator.
Who is it for?
Claude Code developers who finish long debugging sessions and want audited, reusable skills instead of losing fixes in chat history.
Skip if: Developers on Cursor or Codex without Claude Code Stop hooks, or teams that only need one-off answers with no cross-session memory.
When should I use this skill?
A Claude Code session ends after substantial debugging and the developer wants reusable patterns extracted automatically.
What you get
Reviewable learned SKILL.md files under ~/.claude/skills/learned/ capturing error fixes, workarounds, and project-specific techniques.
- learned SKILL.md files
- session evaluation logs
By the numbers
- Detects 5 pattern types including error_resolution and debugging_techniques
- Ignores 3 noise categories such as simple_typos and one_time_fixes
- Default min_session_length is 10 messages
Files
cc-skill-continuous-learning
Development skill skill.
When to Use
This skill is applicable to execute the workflow or actions described in the overview.
Limitations
- Use this skill only when the task clearly matches the scope described above.
- Do not treat the output as a substitute for environment-specific validation, testing, or expert review.
- Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.
{
"min_session_length": 10,
"extraction_threshold": "medium",
"auto_approve": false,
"learned_skills_path": "~/.claude/skills/learned/",
"patterns_to_detect": [
"error_resolution",
"user_corrections",
"workarounds",
"debugging_techniques",
"project_specific"
],
"ignore_patterns": [
"simple_typos",
"one_time_fixes",
"external_api_issues"
]
}
#!/bin/bash
# Continuous Learning - Session Evaluator
# Runs on Stop hook to extract reusable patterns from Claude Code sessions
#
# Why Stop hook instead of UserPromptSubmit:
# - Stop runs once at session end (lightweight)
# - UserPromptSubmit runs every message (heavy, adds latency)
#
# Hook config (in ~/.claude/settings.json):
# {
# "hooks": {
# "Stop": [{
# "matcher": "*",
# "hooks": [{
# "type": "command",
# "command": "~/.claude/skills/continuous-learning/evaluate-session.sh"
# }]
# }]
# }
# }
#
# Patterns to detect: error_resolution, debugging_techniques, workarounds, project_specific
# Patterns to ignore: simple_typos, one_time_fixes, external_api_issues
# Extracted skills saved to: ~/.claude/skills/learned/
set -e
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
CONFIG_FILE="$SCRIPT_DIR/config.json"
LEARNED_SKILLS_PATH="${HOME}/.claude/skills/learned"
MIN_SESSION_LENGTH=10
# Load config if exists
if [ -f "$CONFIG_FILE" ]; then
MIN_SESSION_LENGTH=$(jq -r '.min_session_length // 10' "$CONFIG_FILE")
LEARNED_SKILLS_PATH=$(jq -r '.learned_skills_path // "~/.claude/skills/learned/"' "$CONFIG_FILE" | sed "s|~|$HOME|")
fi
# Ensure learned skills directory exists
mkdir -p "$LEARNED_SKILLS_PATH"
# Get transcript path from environment (set by Claude Code)
transcript_path="${CLAUDE_TRANSCRIPT_PATH:-}"
if [ -z "$transcript_path" ] || [ ! -f "$transcript_path" ]; then
exit 0
fi
# Count messages in session
message_count=$(grep -c '"type":"user"' "$transcript_path" 2>/dev/null || echo "0")
# Skip short sessions
if [ "$message_count" -lt "$MIN_SESSION_LENGTH" ]; then
echo "[ContinuousLearning] Session too short ($message_count messages), skipping" >&2
exit 0
fi
# Signal to Claude that session should be evaluated for extractable patterns
echo "[ContinuousLearning] Session has $message_count messages - evaluate for extractable patterns" >&2
echo "[ContinuousLearning] Save learned skills to: $LEARNED_SKILLS_PATH" >&2
Related skills
How it compares
Pick this over manual skill authoring when Claude Code sessions already contain the debugging narrative and you only need automated extraction at session end.
FAQ
When does cc-skill-continuous-learning run?
cc-skill-continuous-learning runs on the Claude Code Stop hook once at session end, not on UserPromptSubmit. The evaluator script fires after the conversation stops, keeping per-message latency unchanged while still scanning the full transcript.
Where are extracted patterns saved?
cc-skill-continuous-learning writes approved patterns to ~/.claude/skills/learned/ as reusable SKILL.md files. The default learned_skills_path in config points there, and auto_approve is false so nothing lands without review.
What session length triggers extraction?
cc-skill-continuous-learning requires min_session_length of 10 messages before the Stop-hook evaluator attempts extraction. Shorter chats are skipped to avoid promoting trivial or incomplete debugging notes.
Is Cc Skill Continuous Learning safe to install?
skills.sh reports 2 of 3 security scanners passed. Review the Security Audits panel on this page before installing in production.