Settings Reference

Settings Reference

Complete reference for every DeepLore Enhanced setting, organized by section.

Connection

Setting	Default	Range	Description
Enable DeepLore Enhanced	Off	Toggle	Master toggle. When disabled, no entries are injected.

Vault Connections

DeepLore Enhanced supports multiple Obsidian vaults. Each vault has its own name, port, API key, and enable toggle. Entries from all enabled vaults are merged into a single index.

Setting	Default	Description
Vault Name	(none)	Display name for this vault connection. Must match your Obsidian vault name exactly for deep links to work.
Host	127.0.0.1	IP or hostname of the machine running Obsidian. Change for remote vault connections.
Port	27123	Port for the Obsidian Local REST API plugin.
API Key	(none)	Bearer token from Obsidian's Local REST API settings.
Enabled	On	Toggle this vault on/off without deleting the connection.

| Multi-Vault Conflict Resolution | all | Dropdown | How to handle entries with the same title across vaults. all: keep all (disambiguated by vault source). first: keep the first vault's version. last: keep the last vault's version. merge: merge content from all vaults. |

Add Vault button adds a new vault connection. Test All button verifies all enabled vault connections.

Vault Tags

Setting	Default	Description
Lorebook Tag	lorebook	Obsidian tag (without #) that marks a note as a lorebook entry. Only notes with this tag are indexed.
Always-Send Tag	lorebook-always	Tag that forces a note to always be injected regardless of keyword matches. Like constant: true.
Never-Insert Tag	lorebook-never	Tag that prevents a note from ever being injected, even if keywords match. Good for drafts or WIP notes.
Seed Tag	lorebook-seed	Tag for entries whose content is sent to the AI as story context on new chats. Not injected; only informs AI selection. See Features#New Chat Features.
Bootstrap Tag	lorebook-bootstrap	Tag for entries that force-inject when chat is short, then become regular entries. See Features#New Chat Features.
New Chat Threshold	3	1-20. Message count below which seed context is sent and bootstrap entries are force-injected.

Search Mode

Setting	Default	Description
Search Mode	Keyword Only	Dropdown. Keyword Only: keywords-only matching (no AI). Two-Stage: keywords pre-filter, then AI selects. AI Only: entire vault sent to AI. See AI Search.

Matching & Budget

Setting	Default	Range	Description
Scan Depth	4	0-100	Number of recent chat messages to scan for keyword matches. Set to 0 to disable keyword matching (AI search only).
Case Sensitive	Off	Toggle	When on, keyword matching respects case (Eris won't match eris).
Match Whole Words	Off	Toggle	When on, keywords use word boundaries (war won't match warning).
Active Character Boost	Off	Toggle	Auto-match the active character's vault entry by name or keyword, even if not mentioned in chat. See Features#Active Character Boost.
Fuzzy Search (BM25)	Off	Toggle	Supplement keyword matching with BM25/TF-IDF scoring. Helps find entries with partial or approximate keyword matches. Built during index build.
Fuzzy Min Score	0.5	0.1-2.0	(Shown when Fuzzy Search is on.) Minimum BM25 score for a fuzzy match to count. Lower = more permissive, higher = stricter.
Recursive Scanning	Off	Toggle	After initial matches, scan matched entries' content for keywords that trigger more entries.
Max Recursion Steps	3	1-10	Maximum recursive scan passes. Each pass scans newly matched entries for more triggers.
Re-injection Cooldown	0	0-50	Skip re-injecting an entry for N generations after last injection. 0 = disabled. Constants are exempt.
Optimize Keys Mode	keyword-only	Dropdown	Controls keyword optimization strategy. keyword-only uses only the entry's defined keys; two-stage uses keys plus content analysis for better matching.
Strip Duplicate Injections	On	Toggle	Skip re-injecting entries that were already injected in recent generations. Tracked per-chat. Constants are exempt.
Lookback Depth	2	1-10	Number of previous generations to check for already-injected entries (when Strip Duplicate Injections is on). Higher = more aggressive deduplication.
Unlimited Entries	Off	Toggle	Remove the cap on how many entries can be injected per generation.
Max Entries	10	1-100	Maximum entries to inject (when Unlimited Entries is off). Sorted by priority.
Unlimited Token Budget	Off	Toggle	Remove the token budget cap. A warning toast appears if injected lore exceeds 20% of context.
Token Budget	3072	100-100000	Maximum total tokens to inject (when Unlimited Token Budget is off). Entries added in priority order until budget is reached.
Keyword Occurrence Weighting	Off	Toggle	When on, entries with more keyword occurrences in the scan text are weighted higher during matching. Experimental.
Contextual Gating Tolerance	strict	Dropdown	How strictly contextual gating filters entries. strict: entry must match all set filters exactly. moderate: partial matches allowed. lenient: only blocks on direct conflicts. Applies to all gating fields (built-in and custom). Custom fields can also have per-field tolerance overrides.
Field Definitions Path	DeepLore/field-definitions.yaml	Text	Vault path to the YAML file containing custom gating field definitions. Edit Fields button opens the visual rule builder where you can add, remove, or modify custom gating fields (type, operator, tolerance). The four default fields (era, location, scene_type, character_present) are configurable defaults that can be modified or removed.

Injection

Setting	Default	Description
Injection Mode	Extension Prompt	Extension Prompt: uses setExtensionPrompt() with fixed position/depth/role (classic behavior). Prompt List: registers named prompts (deeplore_constants, deeplore_lore) that appear in SillyTavern's Prompt Manager list. Drag them wherever you want. Requires Chat Completion API.
Injection Template	<{{title}}>\n{{content}}\n</{{title}}>	Format for each injected entry. Use {{title}} for entry name and {{content}} for note body.
Injection Position	In-chat	(Extension Prompt mode only) Where to inject lore. Options: Before Main Prompt/Story String, After Main Prompt/Story String, or In-chat @ Depth.
Injection Depth	4	0-9999. Chat depth for in-chat injection (0 = last message).
Injection Role	System	Message role for in-chat injection: System, User, or Assistant.
Allow World Info Scan	Off	Let SillyTavern's built-in World Info system scan injected lore for WI keyword matches. Enables cross-system triggering.

Entries can override position, depth, and role via frontmatter. See Writing Vault Entries.

In Prompt List mode, global position/depth/role settings are ignored — the Prompt Manager controls placement. Per-entry frontmatter overrides with custom depth/position still create separate injection groups that bypass the PM.

Context Cartographer

Setting	Default	Description
Show Lore Sources Button	On	Add a book icon to AI messages showing which entries were injected and why. See Features#Context Cartographer.

Deep links use the vault connection name to build Obsidian URIs. Set vault names in Vault Connections to match your Obsidian vault names exactly.

Author's Notebook

Setting	Default	Description
Enable Author's Notebook	Off	Enable a persistent per-chat scratchpad that is injected into every generation. Edit via /dle-notebook or the Open Notebook button. See Features#Author's Notebook.
Notebook Injection Position	In-chat	Where to inject the notebook. Same options as main injection: Before Main Prompt, After Main Prompt, or In-chat @ Depth.
Notebook Injection Depth	4	0-9999. Chat depth for in-chat notebook injection.
Notebook Injection Role	System	Message role for in-chat notebook injection: System, User, or Assistant.

Open Notebook button opens the notebook editor for the current chat.

AI Search

Visible when Search Mode is Two-Stage or AI Only.

Setting	Default	Range	Description
Connection	Profile	Toggle	"Connection Profile" uses a saved ST profile (recommended). "Custom Proxy" uses a separate proxy server.
Connection Profile	(none)	Dropdown	(Profile mode) Select a saved Connection Manager profile. Any provider works.
Proxy URL	http://localhost:42069	Text	(Proxy mode) URL of the claude-code-proxy or compatible endpoint. Must expose /v1/messages.
Model Override	(none)	Text	Optional model override. In profile mode, leave empty to use the profile's model. In proxy mode, specify the model name.
Max Response Tokens	1024	64-4096	Token limit for the AI response. Keep low; only a JSON array is needed.
Timeout (ms)	10000	1000-120000	How long to wait for the AI before falling back to keyword-only results. Local LLMs may need 60000-120000ms; cloud APIs typically respond in 5000-15000ms.
AI Scan Depth	4	1-100	Number of recent messages to send as context to the AI. Can differ from keyword scan depth.
Entry Description Length	600	100-1000	Max characters for entry descriptions in the AI manifest. Only for entries without a summary field.
System Prompt Override	(none)	Text	Custom system prompt for AI selection. Leave empty for default. Supports {{maxEntries}} placeholder.
Prepend "You are Claude Code"	On	Toggle	(Proxy mode only, under Show Advanced) Prepend You are Claude Code. to the AI system prompt. Disable if using a non-Claude model via proxy.
Use Session Notes as AI Context	Off	Toggle	Feed the Session Scribe's latest summary into the AI search context for better entry selection. See Features#Use Session Notes as AI Context.
AI Confidence Threshold	low	Dropdown	Minimum confidence level for AI selections. low: accept all (high+medium+low). medium: accept medium and high only. high: accept high confidence only.
Hierarchical Pre-filter Aggressiveness	0.8	0.0-0.8	How aggressively the hierarchical category pre-filter culls entries. 0.0 = keep all categories. 0.8 = aggressive filtering. The safety valve kicks in when filtering would remove more than this fraction of entries.
Manifest Summary Mode	prefer_summary	Dropdown	How entry descriptions are built for the AI manifest. prefer_summary: use summary field if present, fall back to truncated content. summary_only: only include entries with summaries. content_only: always use truncated content, ignore summaries.
AI Error Fallback	keyword	Dropdown	What happens when the AI returns an error or times out. keyword: fall back to keyword-matched results. constants_only: inject only constants. bootstrap_only: inject constants and bootstrap entries. none: inject nothing.
AI Empty Result Fallback	constants	Dropdown	What happens when the AI intentionally returns []. constants: inject only constants. constants_bootstrap: inject constants and bootstrap entries. keyword: fall back to keyword results. none: inject nothing.

Test AI Search button tests the AI connection. Preview AI Prompt button shows the full prompt that would be sent.

AI Stats shows session usage: AI calls, cache hits, estimated input/output tokens.

AI-Powered Features

Session Scribe and Auto Lorebook are grouped under one collapsible drawer in settings.

Session Scribe

Setting	Default	Range	Description
Enable Session Scribe	Off	Toggle	Auto-summarize sessions to your Obsidian vault. See Features#Session Scribe.
Auto-Scribe Every N Messages	5	1-50	Number of new messages (tracked by chat position) between automatic summaries.
Session Folder	Sessions	Text	Vault folder where session notes are saved. Created if it doesn't exist.
Connection	SillyTavern	Radio	SillyTavern (active connection), Connection Profile (saved profile), or Custom Proxy.
Connection Profile	(none)	Select	Connection Manager profile for summaries. Only shown in Profile mode.
Proxy URL	http://localhost:42069	Text	Proxy server URL. Only shown in Proxy mode.
Model Override	(none)	Text	Override the model used for summaries. Only shown in Profile/Proxy modes.
Max Response Tokens	1024	256-4096	Maximum tokens for the summary response. Only shown in Profile/Proxy modes.
Timeout (ms)	60000	5000-120000	Request timeout for summary generation. Only shown in Profile/Proxy modes.
Messages to Include	20	5-100	Number of recent chat messages included as context for the summary.
Custom Summary Prompt	(none)	Text	Override the default summary prompt. Default covers events, character dynamics, revelations, and state changes.

Auto Lorebook

Setting	Default	Range	Description
Enable Auto Lorebook	Off	Toggle	AI analyzes chat for entities not in the lorebook and suggests new entries with human review. See Features#Auto Lorebook.
Interval (messages)	10	3-50	Trigger auto-suggest every N messages.
Target Folder	(none)	Text	Obsidian folder for new entries. Leave empty for vault root.
Connection Mode	SillyTavern	Radio	SillyTavern (active connection), Profile (saved profile), or Proxy.
Connection Profile	(none)	Select	Connection Manager profile. Only shown in Profile mode.
Proxy URL	http://localhost:42069	Text	Proxy server URL. Only shown in Proxy mode.
Model	(none)	Text	Model override for suggestions.
Max Tokens	2048	256-4096	Maximum tokens for the suggestion response.
Timeout (ms)	30000	5000-120000	Request timeout for auto-suggest generation. Local LLMs may need 60000-120000ms.
Skip Review	Off	Toggle	When on, auto-suggested entries are written to the vault immediately without showing the review popup. Use with caution.

Use /dle-newlore to trigger on-demand at any time.

Entry Decay

Setting	Default	Range	Description
Enable Entry Decay	Off	Toggle	Track entry freshness and adjust AI manifest priorities. Stale entries get a boost; frequently injected entries get a penalty. See Features#Entry Decay & Freshness.
Mark Stale After N Skips	5	2-20	Consecutive generations an entry is skipped before it gets a freshness boost in the AI manifest.
Mark Frequent After N Injections	2	2-10	Consecutive generations an entry is injected before it gets a frequency penalty in the AI manifest.

Index & Cache

Setting	Default	Range	Description
Cache Duration (seconds)	300	0-86400	How long to cache the vault index before re-fetching. 0 = always fetch fresh (slower).
Auto-Sync Interval	0	0-3600	Seconds between automatic vault re-checks. 0 = disabled. See Features#Vault Change Detection & Auto-Sync.
Show Sync Change Toasts	On	Toggle	Show toast notifications when vault changes are detected during index refresh.
Index Rebuild Trigger	ttl	Dropdown	When to rebuild the vault index. ttl: rebuild when cache duration expires (default). generation: rebuild every N generations. manual: only rebuild on explicit refresh.
Rebuild Every N Generations	10	1-100	(Shown when trigger is generation.) How many generations between automatic index rebuilds.

Refresh Index button clears the cache and re-fetches all entries. Test Match button simulates a generation to show which entries would match.

Show Advanced toggles: Several sections have "Show Advanced" toggles that reveal power-user settings. These toggles persist across sessions. Settings behind advanced toggles include: seed/bootstrap tags, case sensitivity, whole word matching, fuzzy search (BM25), fuzzy min score, recursive scanning, re-injection cooldown, optimize keys mode, deduplication, keyword occurrence weighting, contextual gating tolerance, injection template, WI scan, AI confidence threshold, hierarchical aggressiveness, manifest summary mode, AI error/empty fallbacks, system prompt override, Claude Code prefix, cache TTL, sync interval, sync toasts, index rebuild trigger, and rebuild generation interval.

Advanced

Setting	Default	Range	Description
Review Response Tokens	0	0-100000	Token limit for /dle-review responses. 0 = auto (uses your connection profile's Max Response Length).
Debug Mode	Off	Toggle	Log detailed match info to browser console (F12). Shows keyword matches, AI results, gating, token counts, injection details.

Automatic Features (No Settings)

These features work automatically with no configuration:

• Obsidian Circuit Breaker: Obsidian connection uses a per-vault circuit breaker (closed/open/half-open) with exponential backoff (2s-15s). Keyed by host:port — each vault has independent failure tracking. Prevents hammering a down server. Resets automatically when a call succeeds.
• AI Circuit Breaker: AI search has its own circuit breaker (2 consecutive failures to trip, 30s cooldown). Prevents repeated full-timeout waits when the AI service is down. The AI throttle (2s minimum between calls) does NOT trip the circuit breaker.
• IndexedDB Persistent Cache: Parsed vault index is saved to IndexedDB after every successful build. On page load, hydrates instantly from cache, then validates in background. No settings to configure.
• Reuse Sync: Auto-sync fetches all file contents but skips re-parsing/tokenizing unchanged entries (detected by content hash). Falls back to full rebuild automatically.
• Hierarchical Manifest Clustering: For vaults with 40+ entries and 4+ categories, automatically uses two-call AI approach for better scaling.
• Sliding Window AI Cache: AI search cache tracks manifest and chat hashes separately for smarter cache reuse.
• Confidence-Gated Budget: AI search over-requests entries (2x), sorts by confidence tier before budget cap.
• Prompt Cache Optimization: In proxy mode, manifest is placed first with cache_control breakpoints for Anthropic prompt caching.