fix: prevent /new panic from SQLite pool contention

[BnjmnZmmrmn-parafin]

Summary

Fix a panic on /new caused by SQLite lock contention when ForgeRepo created a second connection pool pointing at the same .forge.db file while the previous pool was still alive.

Context

Every time a user ran /new, ForgeRepo::new created a brand-new DatabasePool via DatabasePool::try_from(...).unwrap(). The old pool — still held alive by tokio::spawn'd hydration tasks — would hold open connections to the same SQLite file. r2d2's pool builder validates connections at construction time with a 5-second timeout; after 5 retries the error bubbled back to the .unwrap(), crashing the process with:

ERROR: called `Result::unwrap()` on an `Err` value: Failed to create connection pool: timed out waiting for connection

Changes

• forge_repo: ForgeRepo::new now accepts an Arc<DatabasePool> instead of constructing its own. DatabasePool and PoolConfig are re-exported from the crate root so callers don't need to reach into internal modules.
• forge_api: ForgeAPI::init creates the pool once and passes it to ForgeRepo::new. Return type changes from Self to Result<Self> — pool failure is now a clean error instead of a panic.
• forge_main: UI's factory closure bound changes from Fn(ForgeConfig) -> A to Fn(ForgeConfig) -> Result<A>. on_new propagates the error via ? instead of unwrapping.

Key Implementation Details

The pool is created once per forge process inside ForgeAPI::init using ConfigReader::base_path().join(".forge.db") — the same canonical path used by Environment::database_path(). On each /new, a new ForgeRepo is wired to the same Arc<DatabasePool>, so the underlying SQLite file is only ever opened once and there is no contention window.

Testing

cargo build -p forge_repo -p forge_api -p forge_main

# Manually verify: start forge, run /new multiple times, confirm no crash
forge
/new
/new
/new

[BnjmnZmmrmn-parafin]

verified by building locally, openning session and spamming /new repeatedly, seemed to work fine

[autofix-ci]

Hi! I'm autofix.ci, a bot that automatically fixes trivial issues such as code formatting in pull requests.

I would like to apply some automated changes to this pull request, but it looks like I don't have the necessary permissions to do so. To get this pull request into a mergeable state, please do one of the following two things:

1. Allow edits by maintainers for your pull request, and then re-trigger CI (for example by pushing a new commit).
2. Manually fix the issues identified for your pull request (see the GitHub Actions output for details on what I would like to change).

[BnjmnZmmrmn-parafin]

Verification

Tested locally against the fix branch using a debug build with FORGE_LOG=debug.

Steps:

1. Built the fix branch: cargo build -p forge_main
2. Ran with debug logging: FORGE_LOG=debug ./target/debug/forge
3. Ran /new repeatedly (10+ times back to back) in the forge prompt
4. Tailed the log file at ~/forge/logs/forge.log.<date> to observe pool lifecycle

What we observed:

Each /new produced a clean Creating database pool → created connection pool pair, completing in ~1ms with zero retries and zero contention errors:

{"timestamp":"   6.223s","level":"DEBUG","fields":{"message":"Creating database pool","database_path":"/Users/benjaminzimmerman/forge/.forge.db"}}
{"timestamp":"   6.224s","level":"DEBUG","fields":{"message":"created connection pool","database_path":"/Users/benjaminzimmerman/forge/.forge.db"}}
{"timestamp":"   7.236s","level":"DEBUG","fields":{"message":"Creating database pool","database_path":"/Users/benjaminzimmerman/forge/.forge.db"}}
{"timestamp":"   7.237s","level":"DEBUG","fields":{"message":"created connection pool","database_path":"/Users/benjaminzimmerman/forge/.forge.db"}}
{"timestamp":"   8.212s","level":"DEBUG","fields":{"message":"Creating database pool","database_path":"/Users/benjaminzimmerman/forge/.forge.db"}}
{"timestamp":"   8.212s","level":"DEBUG","fields":{"message":"created connection pool","database_path":"/Users/benjaminzimmerman/forge/.forge.db"}}
{"timestamp":"   9.153s","level":"DEBUG","fields":{"message":"Creating database pool","database_path":"/Users/benjaminzimmerman/forge/.forge.db"}}
{"timestamp":"   9.154s","level":"DEBUG","fields":{"message":"created connection pool","database_path":"/Users/benjaminzimmerman/forge/.forge.db"}}

Pool reinitializes cleanly on every /new with no contention, no retries, and no crash. Pre-fix this would panic on the 2nd or 3rd /new with Failed to create connection pool: timed out waiting for connection.