Teknium
370 lines
16 KiB
Python
370 lines
16 KiB
Python
"""hermes-memory-store — holographic memory plugin using MemoryProvider interface.
|
|
|
|
Registers as a MemoryProvider plugin, giving the agent structured fact storage
|
|
with entity resolution, trust scoring, and HRR-based compositional retrieval.
|
|
|
|
Original plugin by dusterbloom (PR #2351), adapted to the MemoryProvider ABC.
|
|
|
|
Config in $HERMES_HOME/config.yaml (profile-scoped):
|
|
plugins:
|
|
hermes-memory-store:
|
|
db_path: $HERMES_HOME/memory_store.db # omit to use the default
|
|
auto_extract: false
|
|
default_trust: 0.5
|
|
min_trust_threshold: 0.3
|
|
temporal_decay_half_life: 0
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import json
|
|
import logging
|
|
import re
|
|
from pathlib import Path
|
|
from typing import Any, Dict, List
|
|
|
|
from agent.memory_provider import MemoryProvider
|
|
from tools.registry import tool_error
|
|
from utils import is_truthy_value
|
|
from .store import MemoryStore
|
|
from .retrieval import FactRetriever
|
|
from hermes_cli.config import cfg_get
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
|
|
FACT_STORE_SCHEMA = {
|
|
"name": "fact_store",
|
|
"description": (
|
|
"Deep structured memory with algebraic reasoning. "
|
|
"Use alongside the memory tool — memory for always-on context, "
|
|
"fact_store for deep recall and compositional queries.\n\n"
|
|
"ACTIONS (simple → powerful):\n"
|
|
"• add — Store a fact the user would expect you to remember.\n"
|
|
"• search — Keyword lookup ('editor config', 'deploy process').\n"
|
|
"• probe — Entity recall: ALL facts about a person/thing.\n"
|
|
"• related — What connects to an entity? Structural adjacency.\n"
|
|
"• reason — Compositional: facts connected to MULTIPLE entities simultaneously.\n"
|
|
"• contradict — Memory hygiene: find facts making conflicting claims.\n"
|
|
"• update/remove/list — CRUD operations.\n\n"
|
|
"IMPORTANT: Before answering questions about the user, ALWAYS probe or reason first."
|
|
),
|
|
"parameters": {
|
|
"type": "object",
|
|
"properties": {
|
|
"action": {
|
|
"type": "string",
|
|
"enum": ["add", "search", "probe", "related", "reason", "contradict", "update", "remove", "list"],
|
|
},
|
|
"content": {"type": "string", "description": "Fact content (required for 'add')."},
|
|
"query": {"type": "string", "description": "Search query (required for 'search')."},
|
|
"entity": {"type": "string", "description": "Entity name for 'probe'/'related'."},
|
|
"entities": {"type": "array", "items": {"type": "string"}, "description": "Entity names for 'reason'."},
|
|
"fact_id": {"type": "integer", "description": "Fact ID for 'update'/'remove'."},
|
|
"category": {"type": "string", "enum": ["user_pref", "project", "tool", "general"]},
|
|
"tags": {"type": "string", "description": "Comma-separated tags."},
|
|
"trust_delta": {"type": "number", "description": "Trust adjustment for 'update'."},
|
|
"min_trust": {"type": "number", "description": "Minimum trust filter (default: 0.3)."},
|
|
"limit": {"type": "integer", "description": "Max results (default: 10)."},
|
|
},
|
|
"required": ["action"],
|
|
},
|
|
}
|
|
|
|
FACT_FEEDBACK_SCHEMA = {
|
|
"name": "fact_feedback",
|
|
"description": (
|
|
"Rate a fact after using it. Mark 'helpful' if accurate, 'unhelpful' if outdated. "
|
|
"This trains the memory — good facts rise, bad facts sink."
|
|
),
|
|
"parameters": {
|
|
"type": "object",
|
|
"properties": {
|
|
"action": {"type": "string", "enum": ["helpful", "unhelpful"]},
|
|
"fact_id": {"type": "integer", "description": "The fact ID to rate."},
|
|
},
|
|
"required": ["action", "fact_id"],
|
|
},
|
|
}
|
|
|
|
# Auto-extraction patterns (on_session_end): user preferences -> user_pref, decisions -> project.
|
|
_PREF_PATTERNS = [
|
|
re.compile(r'\bI\s+(?:prefer|like|love|use|want|need)\s+(.+)', re.IGNORECASE),
|
|
re.compile(r'\bmy\s+(?:favorite|preferred|default)\s+\w+\s+is\s+(.+)', re.IGNORECASE),
|
|
re.compile(r'\bI\s+(?:always|never|usually)\s+(.+)', re.IGNORECASE),
|
|
]
|
|
_DECISION_PATTERNS = [
|
|
re.compile(r'\bwe\s+(?:decided|agreed|chose)\s+(?:to\s+)?(.+)', re.IGNORECASE),
|
|
re.compile(r'\bthe\s+project\s+(?:uses|needs|requires)\s+(.+)', re.IGNORECASE),
|
|
]
|
|
|
|
|
|
def _load_plugin_config() -> dict:
|
|
try:
|
|
# Canonical loader: honors the managed-scope overlay + ${VAR} expansion.
|
|
from hermes_cli.config import load_config_readonly
|
|
all_config = load_config_readonly()
|
|
return cfg_get(all_config, "plugins", "hermes-memory-store", default={}) or {}
|
|
except Exception:
|
|
return {}
|
|
|
|
|
|
def _results(results: list) -> str:
|
|
return json.dumps({"results": results, "count": len(results)})
|
|
|
|
|
|
class HolographicMemoryProvider(MemoryProvider):
|
|
"""Holographic memory with structured facts, entity resolution, and HRR retrieval."""
|
|
|
|
def __init__(self, config: dict | None = None):
|
|
self._config = config or _load_plugin_config()
|
|
self._store = None
|
|
self._retriever = None
|
|
self._min_trust = float(self._config.get("min_trust_threshold", 0.3))
|
|
|
|
@property
|
|
def name(self) -> str:
|
|
return "holographic"
|
|
|
|
def is_available(self) -> bool:
|
|
return True # SQLite is always available, numpy is optional
|
|
|
|
def save_config(self, values, hermes_home):
|
|
"""Write config to config.yaml under plugins.hermes-memory-store."""
|
|
config_path = Path(hermes_home) / "config.yaml"
|
|
try:
|
|
import yaml
|
|
# Raw read for the write-back round-trip: merged defaults must not
|
|
# be persisted into the user's file.
|
|
from hermes_cli.config import read_user_config_raw
|
|
existing = read_user_config_raw(config_path)
|
|
existing.setdefault("plugins", {})["hermes-memory-store"] = values
|
|
with open(config_path, "w", encoding="utf-8") as f:
|
|
yaml.dump(existing, f, default_flow_style=False)
|
|
except Exception:
|
|
pass
|
|
|
|
def get_config_schema(self):
|
|
from hermes_constants import display_hermes_home
|
|
_default_db = f"{display_hermes_home()}/memory_store.db"
|
|
return [
|
|
{"key": "db_path", "description": "SQLite database path", "default": _default_db},
|
|
{"key": "auto_extract", "description": "Auto-extract facts at session end", "default": "false", "choices": ["true", "false"]},
|
|
{"key": "default_trust", "description": "Default trust score for new facts", "default": "0.5"},
|
|
{"key": "hrr_dim", "description": "HRR vector dimensions", "default": "1024"},
|
|
]
|
|
|
|
def initialize(self, session_id: str, **kwargs) -> None:
|
|
from hermes_constants import get_hermes_home
|
|
_hermes_home = str(get_hermes_home())
|
|
db_path = self._config.get("db_path", _hermes_home + "/memory_store.db")
|
|
# Expand $HERMES_HOME so configured paths resolve to the active profile's directory.
|
|
if isinstance(db_path, str):
|
|
db_path = db_path.replace("$HERMES_HOME", _hermes_home).replace("${HERMES_HOME}", _hermes_home)
|
|
hrr_dim = int(self._config.get("hrr_dim", 1024))
|
|
|
|
self._store = MemoryStore(
|
|
db_path=db_path, default_trust=float(self._config.get("default_trust", 0.5)), hrr_dim=hrr_dim,
|
|
)
|
|
self._retriever = FactRetriever(
|
|
store=self._store,
|
|
temporal_decay_half_life=int(self._config.get("temporal_decay_half_life", 0)),
|
|
hrr_weight=float(self._config.get("hrr_weight", 0.3)),
|
|
hrr_dim=hrr_dim,
|
|
)
|
|
self._session_id = session_id
|
|
|
|
def system_prompt_block(self) -> str:
|
|
if not self._store:
|
|
return ""
|
|
try:
|
|
total = self._store._conn.execute("SELECT COUNT(*) FROM facts").fetchone()[0]
|
|
except Exception:
|
|
total = 0
|
|
if total == 0:
|
|
return (
|
|
"# Holographic Memory\n"
|
|
"Active. Empty fact store — proactively add facts the user would expect you to remember.\n"
|
|
"Use fact_store(action='add') to store durable structured facts about people, projects, preferences, decisions.\n"
|
|
"Use fact_feedback to rate facts after using them (trains trust scores)."
|
|
)
|
|
return (
|
|
f"# Holographic Memory\n"
|
|
f"Active. {total} facts stored with entity resolution and trust scoring.\n"
|
|
f"Use fact_store to search, probe entities, reason across entities, or add facts.\n"
|
|
f"Use fact_feedback to rate facts after using them (trains trust scores)."
|
|
)
|
|
|
|
def prefetch(self, query: str, *, session_id: str = "") -> str:
|
|
if not self._retriever or not query:
|
|
return ""
|
|
try:
|
|
results = self._retriever.search(query, min_trust=self._min_trust, limit=5)
|
|
if not results:
|
|
return ""
|
|
lines = [f"- [{r.get('trust_score', r.get('trust', 0)):.1f}] {r.get('content', '')}" for r in results]
|
|
return "## Holographic Memory\n" + "\n".join(lines)
|
|
except Exception as e:
|
|
logger.debug("Holographic prefetch failed: %s", e)
|
|
return ""
|
|
|
|
def sync_turn(self, user_content: str, assistant_content: str, *, session_id: str = "") -> None:
|
|
# Facts are stored explicitly via tools; on_session_end handles auto-extraction if configured.
|
|
pass
|
|
|
|
def get_tool_schemas(self) -> List[Dict[str, Any]]:
|
|
return [FACT_STORE_SCHEMA, FACT_FEEDBACK_SCHEMA]
|
|
|
|
def handle_tool_call(self, tool_name: str, args: Dict[str, Any], **kwargs) -> str:
|
|
handler = self._TOOL_HANDLERS.get(tool_name)
|
|
if handler is None:
|
|
return tool_error(f"Unknown tool: {tool_name}")
|
|
try:
|
|
return handler(self, args)
|
|
except KeyError as exc:
|
|
return tool_error(f"Missing required argument: {exc}")
|
|
except Exception as exc:
|
|
return tool_error(str(exc))
|
|
|
|
def on_session_end(self, messages: List[Dict[str, Any]]) -> None:
|
|
# is_truthy_value: the config schema declares auto_extract as a string
|
|
# enum ("false"/"true"); plain truthiness would treat "false" as enabled.
|
|
if is_truthy_value(self._config.get("auto_extract", False)) and self._store and messages:
|
|
self._auto_extract_facts(messages)
|
|
|
|
def on_memory_write(self, action: str, target: str, content: str) -> None:
|
|
"""Mirror built-in memory writes as facts."""
|
|
if action == "add" and self._store and content:
|
|
try:
|
|
category = "user_pref" if target == "user" else "general"
|
|
self._store.add_fact(content, category=category)
|
|
except Exception as e:
|
|
logger.debug("Holographic memory_write mirror failed: %s", e)
|
|
|
|
def shutdown(self) -> None:
|
|
# Release the shared SQLite connection on the caller's thread: leaving
|
|
# it to GC keeps the connection (and its write lock) alive on a
|
|
# long-running gateway. close() is idempotent and refcount-guarded.
|
|
if self._store is not None:
|
|
try:
|
|
self._store.close()
|
|
except Exception as e:
|
|
logger.debug("Holographic shutdown close() failed: %s", e)
|
|
self._store = None
|
|
self._retriever = None
|
|
|
|
# -- Tool handlers -------------------------------------------------------
|
|
# KeyError from args[...] / Exception are turned into tool_error by handle_tool_call.
|
|
|
|
def _handle_fact_store(self, args: dict) -> str:
|
|
action = args["action"]
|
|
handler = self._FACT_STORE_ACTIONS.get(action)
|
|
if handler is None:
|
|
return tool_error(f"Unknown action: {action}")
|
|
return handler(self, args)
|
|
|
|
def _act_add(self, args: dict) -> str:
|
|
fact_id = self._store.add_fact(args["content"], category=args.get("category", "general"), tags=args.get("tags", ""))
|
|
return json.dumps({"fact_id": fact_id, "status": "added"})
|
|
|
|
def _act_search(self, args: dict) -> str:
|
|
return _results(self._retriever.search(
|
|
args["query"], category=args.get("category"),
|
|
min_trust=float(args.get("min_trust", self._min_trust)), limit=int(args.get("limit", 10)),
|
|
))
|
|
|
|
def _act_entity(self, args: dict, method: str) -> str:
|
|
"""Shared body of 'probe' and 'related' (single-entity retriever queries)."""
|
|
return _results(getattr(self._retriever, method)(
|
|
args["entity"], category=args.get("category"), limit=int(args.get("limit", 10)),
|
|
))
|
|
|
|
def _act_reason(self, args: dict) -> str:
|
|
entities = args.get("entities", [])
|
|
if not entities:
|
|
return tool_error("reason requires 'entities' list")
|
|
return _results(self._retriever.reason(entities, category=args.get("category"), limit=int(args.get("limit", 10))))
|
|
|
|
def _act_contradict(self, args: dict) -> str:
|
|
return _results(self._retriever.contradict(category=args.get("category"), limit=int(args.get("limit", 10))))
|
|
|
|
def _act_update(self, args: dict) -> str:
|
|
updated = self._store.update_fact(
|
|
int(args["fact_id"]), content=args.get("content"),
|
|
trust_delta=float(args["trust_delta"]) if "trust_delta" in args else None,
|
|
tags=args.get("tags"), category=args.get("category"),
|
|
)
|
|
return json.dumps({"updated": updated})
|
|
|
|
def _act_remove(self, args: dict) -> str:
|
|
return json.dumps({"removed": self._store.remove_fact(int(args["fact_id"]))})
|
|
|
|
def _act_list(self, args: dict) -> str:
|
|
facts = self._store.list_facts(
|
|
category=args.get("category"), min_trust=float(args.get("min_trust", 0.0)), limit=int(args.get("limit", 10)),
|
|
)
|
|
return json.dumps({"facts": facts, "count": len(facts)})
|
|
|
|
def _handle_fact_feedback(self, args: dict) -> str:
|
|
return json.dumps(self._store.record_feedback(int(args["fact_id"]), helpful=args["action"] == "helpful"))
|
|
|
|
_FACT_STORE_ACTIONS = {
|
|
"add": _act_add, "search": _act_search,
|
|
"probe": lambda self, args: self._act_entity(args, "probe"),
|
|
"related": lambda self, args: self._act_entity(args, "related"),
|
|
"reason": _act_reason, "contradict": _act_contradict,
|
|
"update": _act_update, "remove": _act_remove, "list": _act_list,
|
|
}
|
|
_TOOL_HANDLERS = {"fact_store": _handle_fact_store, "fact_feedback": _handle_fact_feedback}
|
|
|
|
# -- Auto-extraction (on_session_end) ------------------------------------
|
|
|
|
def _auto_extract_facts(self, messages: list) -> None:
|
|
# Local import: the compressor module is heavier than this plugin and
|
|
# only needed when auto_extract is on.
|
|
from agent.context_compressor import (
|
|
_MERGED_PRIOR_CONTEXT_HEADER,
|
|
_MERGED_SUMMARY_DELIMITER,
|
|
is_compaction_summary_message,
|
|
)
|
|
|
|
extracted = 0
|
|
for msg in messages:
|
|
if msg.get("role") != "user":
|
|
continue
|
|
content = msg.get("content", "")
|
|
# Compaction handoff summaries arrive as role="user" and reliably
|
|
# match the decision patterns; skip them so the compactor's own
|
|
# output is never stored as a durable fact. A merge-into-tail row
|
|
# holds genuine prior user text BEFORE _MERGED_SUMMARY_DELIMITER
|
|
# (prefixed with the header) and the summary AFTER it — harvest
|
|
# only the pre-delimiter segment.
|
|
if isinstance(content, str) and _MERGED_SUMMARY_DELIMITER in content:
|
|
pre = content.split(_MERGED_SUMMARY_DELIMITER, 1)[0]
|
|
if pre.startswith(_MERGED_PRIOR_CONTEXT_HEADER):
|
|
pre = pre[len(_MERGED_PRIOR_CONTEXT_HEADER):]
|
|
if pre.strip():
|
|
content = pre.strip()
|
|
elif is_compaction_summary_message(msg):
|
|
continue
|
|
elif is_compaction_summary_message(msg):
|
|
continue
|
|
if not isinstance(content, str) or len(content) < 10:
|
|
continue
|
|
|
|
for patterns, category in ((_PREF_PATTERNS, "user_pref"), (_DECISION_PATTERNS, "project")):
|
|
if any(p.search(content) for p in patterns):
|
|
try:
|
|
self._store.add_fact(content[:400], category=category)
|
|
extracted += 1
|
|
except Exception:
|
|
pass
|
|
|
|
if extracted:
|
|
logger.info("Auto-extracted %d facts from conversation", extracted)
|
|
|
|
|
|
def register(ctx) -> None:
|
|
"""Register the holographic memory provider with the plugin system."""
|
|
ctx.register_memory_provider(HolographicMemoryProvider(config=_load_plugin_config()))
|