How Lore Compares

The Landscape

Many tools exist for documenting software decisions, but most share a fundamental flaw: they ask developers to stop what they're doing and write documentation separately from coding. That's like asking a surgeon to write their operation notes the next day from memory.

Lore takes a different approach: capture at the moment of the decision, not after.

Quick Comparison

	Lore	Swimm	Confluence	GitBook	ADRs	Nothing
When	Commit-time	After the fact	After the fact	After the fact	When remembered	Never
Where	Local (`.lore/`)	SaaS	SaaS	SaaS	Local (Markdown)	—
Friction	90 seconds	30 minutes	30 minutes	15 minutes	15 minutes	0
AI	Angela (opt-in)	Generic	Generic AI	Generic AI	None	—
Lock-in	Markdown	Proprietary	Proprietary	Mixed	Markdown	—
Offline	Yes (everything)	No	No	No	Yes	—
Automated	Post-commit hook	Manual	Manual	Manual	Manual	—
Bilingual	EN/FR built-in	EN only	Multi-language	Multi-language	Manual	—
Price	Free (AGPL)	$28/seat	$5.75/user	$8/user	Free	Free

Why Commit-Time Matters

Knowledge has a half-life. At the moment of the commit, the developer knows exactly why they made the choice. One hour later, details start fading. One week later, it's vague. Six months later — gone.

Commit moment  ████████████████████ 100% context
1 hour later   ████████████████░░░░  80% context
1 day later    ████████████░░░░░░░░  60% context
1 week later   ████████░░░░░░░░░░░░  40% context
1 month later  ████░░░░░░░░░░░░░░░░  20% context
6 months later ░░░░░░░░░░░░░░░░░░░░   0% context

Lore captures at the peak. Everything else captures on the decline.

Detailed Comparisons

Lore vs Swimm

Swimm is a SaaS documentation platform that lives alongside your code. It is well-designed and has solid IDE integrations.

Aspect	Lore	Swimm
Capture moment	Automatically at commit	Manually, when you remember
Data location	Your repo (`.lore/docs/`)	Swimm's servers
Offline	Fully functional	Requires internet
AI	Angela: zero-API + optional AI	Generic AI assistant
Price	Free forever	$28/seat/month
Vendor risk	None (Markdown files)	Company could pivot, raise prices, or shut down

When Swimm is better: Large teams that need collaborative editing, visual documentation, and IDE widgets. When Lore is better: Developers who want zero-friction, local-first, commit-time capture with no subscription.

Lore vs Confluence

Confluence is Atlassian's wiki — the default enterprise choice.

The core problem: nobody updates Confluence. Pages rot. The "Authentication Architecture" page was written 18 months ago by someone who has since left. It describes a system that no longer exists. Everyone knows it's wrong, but nobody has time to fix it.

Lore sidesteps this problem because documents are created at the moment of change. They cannot rot silently — lore angela review catches contradictions, and lore doctor flags stale content.

Lore vs ADRs

Architecture Decision Records (ADRs) are Markdown files that document significant architectural decisions. They are great.

Lore is not a replacement for ADRs — they are complementary:

	ADRs	Lore
Scope	Big, rare decisions	Daily commit-level decisions
Frequency	Once a quarter	Every commit
Trigger	Manual ("someone should write an ADR")	Automatic (post-commit hook)
Example	"We chose PostgreSQL over MongoDB"	"Why we added this index to the users table"

The best setup: ADRs for the big picture, Lore for the daily details. Over time, Lore documents naturally feed into ADR discussions.

Lore vs Conventional Commits

Conventional Commits (feat:, fix:, docs:) standardize the what. Lore captures the why. They work beautifully together:

git commit -m "feat(auth): add JWT middleware"
# Conventional Commit tells you: it's a feature, in the auth scope
# Lore asks: WHY JWT? WHY now? What alternatives were considered?

Lore pre-fills the "What" field from your commit message. If you use Conventional Commits, Lore's Decision Engine recognizes the type prefix and adjusts scoring accordingly.

Lore vs Doing Nothing

Most teams do nothing. It works — until it doesn't.

The cost of lost context is real but invisible:

Wrong refactors — Removing code that existed for a reason nobody remembers
Repeated mistakes — Rediscovering the same decision (and same mistake) that was already made and undone
Onboarding friction — New team members spend weeks asking "why is this like this?"
Review delays — PRs stall because reviewers lack the reasoning behind changes

Lore's bet: 90 seconds per commit is worth it. Over a year, that's ~6 hours of documentation for ~1,500 commits. The return: a searchable knowledge base that saves hundreds of hours of "why did we do this?"