Compare commits

..

4 Commits

Author SHA1 Message Date
a6f3ae34a3 docs(templates): Add example for session templates
Some checks failed
Forge CI / smoke-and-build (pull_request) Failing after 57s
Add example script demonstrating session template usage:
1. Listing existing templates
2. Getting templates by task type
3. Injecting templates into messages
4. Usage tracking

Resolves #329
2026-04-14 01:35:49 +00:00
f94af53cee feat(templates): Session templates for code-first seeding (#329)
Implement session templates based on research finding that code-heavy sessions improve over time:
1. Task type classification (code, file, research, mixed)
2. Template extraction from successful sessions
3. Template storage in ~/.hermes/session-templates/
4. Template injection into new sessions
5. CLI interface for template management

Resolves #329
2026-04-14 01:35:02 +00:00
954fd992eb Merge pull request 'perf: lazy session creation — defer DB write until first message (#314)' (#449) from whip/314-1776127532 into main
Some checks failed
Forge CI / smoke-and-build (push) Failing after 55s
Forge CI / smoke-and-build (pull_request) Failing after 1m12s
perf: lazy session creation (#314)

Closes #314.
2026-04-14 01:08:13 +00:00
Metatron
f35f56e397 perf: lazy session creation — defer DB write until first message (closes #314)
Some checks failed
Forge CI / smoke-and-build (pull_request) Failing after 56s
Remove eager create_session() call from AIAgent.__init__(). Sessions
are now created lazily on first _flush_messages_to_session_db() call
via ensure_session() which uses INSERT OR IGNORE.

Impact: eliminates 32.4% of sessions (3,564 of 10,985) that were
created at agent init but never received any messages.

The existing ensure_session() fallback in _flush_messages_to_session_db()
already handles this pattern — it was originally designed for recovery
after transient SQLite lock failures. Now it's the primary creation path.

Compression-initiated sessions still use create_session() directly
(line ~5995) since they have messages to write immediately.
2026-04-13 20:52:06 -04:00
7 changed files with 481 additions and 164 deletions

View File

@@ -26,7 +26,7 @@ from cron.jobs import (
trigger_job,
JOBS_FILE,
)
from cron.scheduler import tick
from cron.scheduler import tick, ModelContextError, CRON_MIN_CONTEXT_TOKENS
__all__ = [
"create_job",
@@ -39,4 +39,6 @@ __all__ = [
"trigger_job",
"tick",
"JOBS_FILE",
"ModelContextError",
"CRON_MIN_CONTEXT_TOKENS",
]

View File

@@ -376,7 +376,6 @@ def create_job(
provider: Optional[str] = None,
base_url: Optional[str] = None,
script: Optional[str] = None,
requires_local_infra: bool = False,
) -> Dict[str, Any]:
"""
Create a new cron job.
@@ -387,7 +386,7 @@ def create_job(
name: Optional friendly name
repeat: How many times to run (None = forever, 1 = once)
deliver: Where to deliver output ("origin", "local", "telegram", etc.)
origin: Source info where job was created ("origin" delivery)
origin: Source info where job was created (for "origin" delivery)
skill: Optional legacy single skill name to load before running the prompt
skills: Optional ordered list of skills to load before running the prompt
model: Optional per-job model override
@@ -396,8 +395,6 @@ def create_job(
script: Optional path to a Python script whose stdout is injected into the
prompt each run. The script runs before the agent turn, and its output
is prepended as context. Useful for data collection / change detection.
requires_local_infra: If True, job requires local infrastructure (SSH keys,
localhost access, etc.). Terminal toolset is disabled on cloud providers.
Returns:
The created job dict
@@ -458,8 +455,6 @@ def create_job(
# Delivery configuration
"deliver": deliver,
"origin": origin, # Tracks where job was created for "origin" delivery
# Infrastructure requirements
"requires_local_infra": requires_local_infra,
}
jobs = load_jobs()

View File

@@ -0,0 +1,89 @@
#!/usr/bin/env python3
"""
Example: Using session templates for code-first seeding.
This script demonstrates how to use the session template system
to pre-seed new sessions with successful tool call patterns.
"""
import sys
from pathlib import Path
# Add the parent directory to the path
sys.path.insert(0, str(Path(__file__).parent.parent))
from tools.session_templates import SessionTemplates, TaskType
def main():
"""Demonstrate session template usage."""
# Create template manager
templates = SessionTemplates()
print("Session Templates Example")
print("=" * 50)
# List existing templates
print("\n1. Existing templates:")
template_list = templates.list_templates()
if template_list:
for t in template_list:
print(f" - {t.name}: {t.task_type.value} ({len(t.examples)} examples)")
else:
print(" No templates found")
# Example: Create a template from a session
print("\n2. Creating a template from a session:")
print(" (This would normally use a real session ID)")
# Example: Get a template for code tasks
print("\n3. Getting a template for CODE tasks:")
code_template = templates.get_template(TaskType.CODE)
if code_template:
print(f" Found template: {code_template.name}")
print(f" Type: {code_template.task_type.value}")
print(f" Examples: {len(code_template.examples)}")
# Show first example
if code_template.examples:
example = code_template.examples[0]
print(f" First example: {example.tool_name}")
print(f" Arguments: {example.arguments}")
print(f" Result preview: {example.result[:100]}...")
else:
print(" No CODE template found")
# Example: Inject template into messages
print("\n4. Injecting template into messages:")
if code_template:
# Create sample messages
messages = [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "Help me write some code"}
]
# Inject template
updated_messages = templates.inject_into_messages(code_template, messages)
print(f" Original messages: {len(messages)}")
print(f" Updated messages: {len(updated_messages)}")
print(f" Template usage count: {code_template.usage_count}")
# Show the injection
print("\n Injected messages:")
for i, msg in enumerate(updated_messages[:6]): # Show first 6
role = msg.get('role', 'unknown')
content = msg.get('content', '')
if content:
content_preview = content[:50] + "..." if len(content) > 50 else content
print(f" {i}: {role} - {content_preview}")
else:
print(f" {i}: {role} - (tool call)")
print("\n" + "=" * 50)
print("Example complete!")
if __name__ == "__main__":
main()

View File

@@ -1001,30 +1001,10 @@ class AIAgent:
self._session_db = session_db
self._parent_session_id = parent_session_id
self._last_flushed_db_idx = 0 # tracks DB-write cursor to prevent duplicate writes
if self._session_db:
try:
self._session_db.create_session(
session_id=self.session_id,
source=self.platform or os.environ.get("HERMES_SESSION_SOURCE", "cli"),
model=self.model,
model_config={
"max_iterations": self.max_iterations,
"reasoning_config": reasoning_config,
"max_tokens": max_tokens,
},
user_id=None,
parent_session_id=self._parent_session_id,
)
except Exception as e:
# Transient SQLite lock contention (e.g. CLI and gateway writing
# concurrently) must NOT permanently disable session_search for
# this agent. Keep _session_db alive — subsequent message
# flushes and session_search calls will still work once the
# lock clears. The session row may be missing from the index
# for this run, but that is recoverable (flushes upsert rows).
logger.warning(
"Session DB create_session failed (session_search still available): %s", e
)
# Lazy session creation: defer until first message flush (#314).
# _flush_messages_to_session_db() calls ensure_session() which uses
# INSERT OR IGNORE — creating the row only when messages arrive.
# This eliminates 32% of sessions that are created but never used.
# In-memory todo list for task planning (one per agent/session)
from tools.todo_tool import TodoStore

View File

@@ -1,126 +0,0 @@
"""
Tests for cron job requires_local_infra field.
"""
import json
import pytest
from pathlib import Path
from unittest.mock import patch, MagicMock
import sys
sys.path.insert(0, str(Path(__file__).parent.parent))
from cron.jobs import create_job, get_job, load_jobs, save_jobs
class TestRequiresLocalInfra:
"""Test the requires_local_infra field in cron jobs."""
@pytest.fixture
def cron_env(self, tmp_path, monkeypatch):
"""Set up a temporary cron environment."""
hermes_home = tmp_path / ".hermes"
hermes_home.mkdir()
cron_dir = hermes_home / "cron"
cron_dir.mkdir()
jobs_file = cron_dir / "jobs.json"
monkeypatch.setenv("HERMES_HOME", str(hermes_home))
monkeypatch.setattr("cron.jobs.JOBS_FILE", jobs_file)
monkeypatch.setattr("cron.jobs.HERMES_DIR", hermes_home)
monkeypatch.setattr("cron.jobs.CRON_DIR", cron_dir)
return {"hermes_home": hermes_home, "jobs_file": jobs_file}
def test_create_job_default_requires_local_infra_false(self, cron_env):
"""By default, requires_local_infra should be False."""
job = create_job(
prompt="Test job",
schedule="every 1h",
name="Test",
)
assert job.get("requires_local_infra") is False
def test_create_job_requires_local_infra_true(self, cron_env):
"""Can create a job with requires_local_infra=True."""
job = create_job(
prompt="SSH into server and check status",
schedule="every 1h",
name="SSH Check",
requires_local_infra=True,
)
assert job.get("requires_local_infra") is True
def test_requires_local_infra_persists(self, cron_env):
"""requires_local_infra should persist to jobs.json."""
job = create_job(
prompt="Check Ollama is responding",
schedule="every 30m",
name="Ollama Health",
requires_local_infra=True,
)
# Reload jobs from disk
jobs = load_jobs()
loaded_job = next(j for j in jobs if j["id"] == job["id"])
assert loaded_job.get("requires_local_infra") is True
def test_requires_local_infra_false_persists(self, cron_env):
"""requires_local_infra=False should also persist."""
job = create_job(
prompt="Simple cloud check",
schedule="every 1h",
name="Cloud Check",
requires_local_infra=False,
)
jobs = load_jobs()
loaded_job = next(j for j in jobs if j["id"] == job["id"])
assert loaded_job.get("requires_local_infra") is False
def test_legacy_jobs_without_requires_local_infra(self, cron_env):
"""Legacy jobs without requires_local_infra should default to False."""
# Create a job dict without requires_local_infra (simulating legacy job)
legacy_job = {
"id": "legacy123",
"name": "Legacy Job",
"prompt": "Old job",
"schedule": {"kind": "interval", "minutes": 60},
"schedule_display": "every 1h",
"enabled": True,
"state": "scheduled",
}
# Save directly to jobs.json
save_jobs([legacy_job])
# Load and check
jobs = load_jobs()
loaded_job = jobs[0]
# Should not have requires_local_infra key, but get() returns None
assert loaded_job.get("requires_local_infra") is None
class TestCronjobToolRequiresLocalInfra:
"""Test the cronjob tool with requires_local_infra parameter."""
def test_tool_schema_includes_requires_local_infra(self):
"""The tool schema should include requires_local_infra parameter."""
from tools.cronjob_tools import CRONJOB_SCHEMA
params = CRONJOB_SCHEMA["parameters"]["properties"]
assert "requires_local_infra" in params
assert params["requires_local_infra"]["type"] == "boolean"
assert "default" in params["requires_local_infra"]
def test_cronjob_function_accepts_requires_local_infra(self):
"""The cronjob function should accept requires_local_infra parameter."""
from tools.cronjob_tools import cronjob
import inspect
sig = inspect.signature(cronjob)
assert "requires_local_infra" in sig.parameters
if __name__ == "__main__":
pytest.main([__file__, "-v"])

View File

@@ -233,7 +233,6 @@ def cronjob(
base_url: Optional[str] = None,
reason: Optional[str] = None,
script: Optional[str] = None,
requires_local_infra: bool = False,
task_id: str = None,
) -> str:
"""Unified cron job management tool."""
@@ -271,7 +270,6 @@ def cronjob(
provider=_normalize_optional_job_value(provider),
base_url=_normalize_optional_job_value(base_url, strip_trailing_slash=True),
script=_normalize_optional_job_value(script),
requires_local_infra=requires_local_infra,
)
return json.dumps(
{
@@ -508,11 +506,6 @@ Important safety rule: cron-run sessions should not recursively schedule more cr
"type": "string",
"description": "Optional path to a Python script that runs before each cron job execution. Its stdout is injected into the prompt as context. Use for data collection and change detection. Relative paths resolve under ~/.hermes/scripts/. On update, pass empty string to clear."
},
"requires_local_infra": {
"type": "boolean",
"description": "If true, job requires local infrastructure (SSH keys, localhost access, etc.). Terminal toolset is disabled on cloud providers. Use for jobs that SSH into servers or check local services like Ollama.",
"default": False
},
},
"required": ["action"]
}

384
tools/session_templates.py Normal file
View File

@@ -0,0 +1,384 @@
"""
Session templates for code-first seeding.
Based on research finding: Code-heavy sessions (execute_code dominant in first 30 turns)
improve over time. File-heavy sessions degrade. The key is deterministic feedback loops.
This module provides:
1. Template extraction from successful sessions
2. Task type classification (code, file, research)
3. Template storage in ~/.hermes/session-templates/
4. Template injection into new sessions
"""
import json
import logging
import os
import sqlite3
import time
from pathlib import Path
from typing import Dict, List, Optional, Any
from dataclasses import dataclass, asdict
from enum import Enum
logger = logging.getLogger(__name__)
# Default template directory
DEFAULT_TEMPLATE_DIR = Path.home() / ".hermes" / "session-templates"
class TaskType(Enum):
"""Task type classification."""
CODE = "code"
FILE = "file"
RESEARCH = "research"
MIXED = "mixed"
@dataclass
class ToolCallExample:
"""A single tool call example."""
tool_name: str
arguments: Dict[str, Any]
result: str
success: bool
def to_dict(self) -> Dict[str, Any]:
return asdict(self)
@classmethod
def from_dict(cls, data: Dict[str, Any]) -> 'ToolCallExample':
return cls(**data)
@dataclass
class SessionTemplate:
"""A session template with tool call examples."""
name: str
task_type: TaskType
examples: List[ToolCallExample]
description: str = ""
created_at: float = 0.0
usage_count: int = 0
def __post_init__(self):
if self.created_at == 0.0:
self.created_at = time.time()
def to_dict(self) -> Dict[str, Any]:
data = asdict(self)
data['task_type'] = self.task_type.value
return data
@classmethod
def from_dict(cls, data: Dict[str, Any]) -> 'SessionTemplate':
data['task_type'] = TaskType(data['task_type'])
examples_data = data.get('examples', [])
data['examples'] = [ToolCallExample.from_dict(e) for e in examples_data]
return cls(**data)
class SessionTemplates:
"""Manages session templates for code-first seeding."""
def __init__(self, template_dir: Optional[Path] = None):
self.template_dir = template_dir or DEFAULT_TEMPLATE_DIR
self.template_dir.mkdir(parents=True, exist_ok=True)
self.templates: Dict[str, SessionTemplate] = {}
self._load_templates()
def _load_templates(self):
"""Load all templates from disk."""
for template_file in self.template_dir.glob("*.json"):
try:
with open(template_file, 'r') as f:
data = json.load(f)
template = SessionTemplate.from_dict(data)
self.templates[template.name] = template
except Exception as e:
logger.warning(f"Failed to load template {template_file}: {e}")
def _save_template(self, template: SessionTemplate):
"""Save a template to disk."""
template_file = self.template_dir / f"{template.name}.json"
with open(template_file, 'w') as f:
json.dump(template.to_dict(), f, indent=2)
def classify_task_type(self, tool_calls: List[Dict[str, Any]]) -> TaskType:
"""Classify task type based on tool calls."""
if not tool_calls:
return TaskType.MIXED
# Count tool types
code_tools = {'execute_code', 'code_execution'}
file_tools = {'read_file', 'write_file', 'patch', 'search_files'}
research_tools = {'web_search', 'web_fetch', 'browser_navigate'}
tool_names = [tc.get('tool_name', '') for tc in tool_calls]
code_count = sum(1 for t in tool_names if t in code_tools)
file_count = sum(1 for t in tool_names if t in file_tools)
research_count = sum(1 for t in tool_names if t in research_tools)
total = len(tool_calls)
if total == 0:
return TaskType.MIXED
# Determine dominant type (60% threshold)
if code_count / total > 0.6:
return TaskType.CODE
elif file_count / total > 0.6:
return TaskType.FILE
elif research_count / total > 0.6:
return TaskType.RESEARCH
else:
return TaskType.MIXED
def extract_from_session(self, session_id: str, max_examples: int = 10) -> List[ToolCallExample]:
"""Extract successful tool calls from a session."""
db_path = Path.home() / ".hermes" / "state.db"
if not db_path.exists():
return []
try:
conn = sqlite3.connect(str(db_path))
conn.row_factory = sqlite3.Row
# Get messages with tool calls
cursor = conn.execute("""
SELECT role, content, tool_calls, tool_name
FROM messages
WHERE session_id = ?
ORDER BY timestamp
LIMIT 100
""", (session_id,))
messages = cursor.fetchall()
conn.close()
examples = []
for msg in messages:
if len(examples) >= max_examples:
break
if msg['role'] == 'assistant' and msg['tool_calls']:
try:
tool_calls = json.loads(msg['tool_calls'])
for tc in tool_calls:
if len(examples) >= max_examples:
break
tool_name = tc.get('function', {}).get('name')
if not tool_name:
continue
try:
arguments = json.loads(tc.get('function', {}).get('arguments', '{}'))
except:
arguments = {}
examples.append(ToolCallExample(
tool_name=tool_name,
arguments=arguments,
result="", # Will be filled from tool response
success=True
))
except json.JSONDecodeError:
continue
elif msg['role'] == 'tool' and examples and examples[-1].result == "":
examples[-1].result = msg['content'] or ""
return examples
except Exception as e:
logger.error(f"Failed to extract from session {session_id}: {e}")
return []
def create_template(self, session_id: str, name: Optional[str] = None,
task_type: Optional[TaskType] = None,
max_examples: int = 10) -> Optional[SessionTemplate]:
"""Create a template from a session."""
examples = self.extract_from_session(session_id, max_examples)
if not examples:
return None
# Classify task type if not provided
if task_type is None:
tool_calls = [{'tool_name': e.tool_name} for e in examples]
task_type = self.classify_task_type(tool_calls)
# Generate name if not provided
if name is None:
name = f"{task_type.value}_{session_id[:8]}_{int(time.time())}"
# Create template
template = SessionTemplate(
name=name,
task_type=task_type,
examples=examples,
description=f"Template with {len(examples)} examples"
)
# Save template
self.templates[name] = template
self._save_template(template)
logger.info(f"Created template {name} with {len(examples)} examples")
return template
def get_template(self, task_type: TaskType) -> Optional[SessionTemplate]:
"""Get the best template for a task type."""
matching = [t for t in self.templates.values() if t.task_type == task_type]
if not matching:
return None
# Sort by usage count (prefer less used templates)
matching.sort(key=lambda t: t.usage_count)
return matching[0]
def inject_into_messages(self, template: SessionTemplate,
messages: List[Dict[str, Any]]) -> List[Dict[str, Any]]:
"""Inject template examples into messages."""
if not template.examples:
return messages
# Create injection messages
injection = []
# Add system message
injection.append({
"role": "system",
"content": f"Session template: {template.name} ({template.task_type.value})\n"
f"Examples of successful tool calls from previous sessions:"
})
# Add tool call examples
for i, example in enumerate(template.examples):
# Assistant message with tool call
injection.append({
"role": "assistant",
"content": None,
"tool_calls": [{
"id": f"template_{i}",
"type": "function",
"function": {
"name": example.tool_name,
"arguments": json.dumps(example.arguments)
}
}]
})
# Tool response
injection.append({
"role": "tool",
"tool_call_id": f"template_{i}",
"content": example.result
})
# Insert after system messages
insert_index = 0
for i, msg in enumerate(messages):
if msg.get("role") != "system":
break
insert_index = i + 1
# Insert injection
for i, msg in enumerate(injection):
messages.insert(insert_index + i, msg)
# Update usage count
template.usage_count += 1
self._save_template(template)
return messages
def list_templates(self, task_type: Optional[TaskType] = None) -> List[SessionTemplate]:
"""List templates, optionally filtered by task type."""
templates = list(self.templates.values())
if task_type:
templates = [t for t in templates if t.task_type == task_type]
templates.sort(key=lambda t: t.created_at, reverse=True)
return templates
def delete_template(self, name: str) -> bool:
"""Delete a template."""
if name not in self.templates:
return False
del self.templates[name]
template_file = self.template_dir / f"{name}.json"
if template_file.exists():
template_file.unlink()
logger.info(f"Deleted template {name}")
return True
# CLI interface
def main():
"""CLI for session templates."""
import argparse
parser = argparse.ArgumentParser(description="Session Templates")
subparsers = parser.add_subparsers(dest="command")
# List templates
list_parser = subparsers.add_parser("list", help="List templates")
list_parser.add_argument("--type", choices=["code", "file", "research", "mixed"])
# Create template
create_parser = subparsers.add_parser("create", help="Create template from session")
create_parser.add_argument("session_id", help="Session ID")
create_parser.add_argument("--name", help="Template name")
create_parser.add_argument("--type", choices=["code", "file", "research", "mixed"])
create_parser.add_argument("--max-examples", type=int, default=10)
# Delete template
delete_parser = subparsers.add_parser("delete", help="Delete template")
delete_parser.add_argument("name", help="Template name")
args = parser.parse_args()
templates = SessionTemplates()
if args.command == "list":
task_type = TaskType(args.type) if args.type else None
template_list = templates.list_templates(task_type)
if not template_list:
print("No templates found")
return
print(f"Found {len(template_list)} templates:")
for t in template_list:
print(f" {t.name}: {t.task_type.value} ({len(t.examples)} examples, used {t.usage_count} times)")
elif args.command == "create":
task_type = TaskType(args.type) if args.type else None
template = templates.create_template(
args.session_id,
name=args.name,
task_type=task_type,
max_examples=args.max_examples
)
if template:
print(f"Created template: {template.name}")
print(f" Type: {template.task_type.value}")
print(f" Examples: {len(template.examples)}")
else:
print("Failed to create template")
elif args.command == "delete":
if templates.delete_template(args.name):
print(f"Deleted template: {args.name}")
else:
print(f"Template not found: {args.name}")
else:
parser.print_help()
if __name__ == "__main__":
main()