@lowercasebryan

[@lowercasebryan @lower_case_b]

@cyrusjohnson But it's on Reddit, bruh!

[@lowercasebryan @lower_case_b]

@rothken PII scrub before sending anything to your org API, zdr is a non-issue.

[@lowercasebryan @lower_case_b]

@ClaudeDevs Also, is the Fable -> Opus fallback still billing at Fable 2x usage/billing? I saw multiple fallbacks and drained ~$1,500 in 1.5 days and sub is just about to hit weekly limit (2 days in!)

[@lowercasebryan @lower_case_b]

@ClaudeDevs It seems to trip Fable -> Opus 4.8 simply because security related or adjacent work is in context from completed tasks. I assume the quick workaround js /clear but would be great if it could distinguish new tasks that is not directly security related.

[@lowercasebryan @lower_case_b]

@BitGrateful @CamoText I have workflows where I scrub during parsing also, depending on how datasets will be stored/accessed.

[Meta Alchemist @meta_alchemist]

Got your hands on Claude Fable 5?

The first thing you should do is to upgrade your main projects with it, so it drastically impoves everything you've been working on.

Run this Audit & Project Improvement Prompt on each repo that's important to you (simply copy-paste it):

Repo Audit & Improvement Plan:
Prompt made by Claude Fable 5

You are a world-class principal-level software engineer and technical auditor. Your job is to deeply analyze this repository, produce an honest audit, and deliver a prioritized, actionable improvement plan. Work in the four phases below, in order. Do not skip ahead.

Ground every claim in actual files: cite file paths and line numbers. If you can't verify something, say so explicitly rather than guessing.

Phase 1 / Discovery & Mapping (read before judging)
Explore the repository systematically before forming any opinions:

Map the directory structure and identify the project type, language(s), frameworks, and runtime targets.
Identify entry points, core modules, and the main data/control flow through the system.

Read the package manifest(s), lockfiles, build config, CI config, environment/config files, and any docs (README, CONTRIBUTING, ADRs).

Determine what the project is for: its purpose, intended users, and apparent maturity (prototype, internal tool, production service, library).

Note conventions already in use (naming, module boundaries, error handling patterns, test style) so recommendations fit the existing culture rather than fighting it.

Output for this phase: a concise "Repo Map" purpose, stack, architecture sketch, key directories with one-line descriptions, and anything that surprised you.

Phase 2 / Audit (evidence-based, severity-rated)

Audit each dimension below.

For every finding, record: (a) what you found, (b) where (file:line), (c) why it matters (concrete consequence, not vague principle), (d) severity:

Critical / High / Medium / Low.

• Architecture & design: module boundaries, coupling/cohesion, circular dependencies, leaky abstractions, god objects/files, layering violations, scalability bottlenecks.

• Code quality: duplication, dead code, complexity hotspots (longest/most-branched functions), inconsistent patterns, error handling gaps (swallowed exceptions, missing edge cases), type safety holes.

• Security: hardcoded secrets or credentials, injection risks, unsafe deserialization, missing input validation, auth/authz weaknesses, outdated dependencies with known CVEs, overly permissive configs.

• Testing: coverage gaps (especially around core business logic), test quality (do tests assert behavior or just execution?), missing test types (unit/integration/e2e), flaky patterns, untestable code.

• Performance: N+1 queries, unnecessary allocations or copies, blocking calls in async paths, missing caching/indexing, unbounded growth (memory, files, queues).

• Dependencies: outdated, unmaintained, duplicated, or unnecessarily heavy packages; license risks; lockfile hygiene.

• DevEx & operations: build/setup friction, CI/CD gaps, missing linting/formatting enforcement, logging/observability quality, error reporting, deployment story.

• Documentation: README accuracy, onboarding path, undocumented critical behavior, stale docs that contradict code.

Rules for this phase:

Prefer 15 high-confidence findings over 50 speculative ones.

Distinguish facts ("this function has no error handling: src/api/client.ts:142") from judgments ("this module's responsibilities feel unclear") and label which is which.

Also list what the repo does well: strengths matter for deciding what to preserve.

Output for this phase: an "Audit Report": findings grouped by dimension, sorted by severity, plus a Strengths section.

Don't forget to mention all the ugly parts that need utmost priority.

Phase 3 / Improvement Strategy

Synthesize the audit into a strategy:

Identify the 3–5 themes that explain most of the findings (e.g., "no enforced boundaries between layers," "error handling is ad hoc").

For each theme, propose a target state and the principle behind it.

State explicit trade-offs: what you're recommending NOT to fix and why (effort vs. payoff, risk, project maturity).

Define what "done" looks like — measurable signals (e.g., "CI fails on lint errors," "core module test coverage ≥ 80%," "zero Critical findings").

Phase 4 / Detailed Task Plan
Convert the strategy into an execution plan:

Break work into discrete tasks. Each task must include: Title and one-paragraph description
Files/areas affected

Acceptance criteria (how we verify it's done)
Effort estimate (S = <2h, M = half-day, L = 1–2 days, XL = needs breakdown)

Risk of the change itself (could it break things?)
Dependencies on other tasks

Order tasks into milestones:

Milestone 0
Safety net: anything needed before refactoring safely (tests around critical paths, CI gates, backups).

Milestone 1
Critical fixes: security and correctness issues.

Milestone 2
High-leverage improvements: changes that make all future work easier.

Milestone 3
Quality & polish: remaining medium/low items worth doing.

Flag quick wins (high impact, S effort) separately so they can be done immediately.

For the top 3 tasks, include a brief implementation sketch (approach, key steps, gotchas).

Final Deliverable Format
• Produce a single document with these sections:
• Executive Summary (≤10 sentences: overall health grade A–F with justification, top 3 risks, top 3 opportunities)
• Repo Map
• Audit Report
• Improvement Strategy
• Task Plan (milestones + task table + quick wins)
• Open Questions: anything you need from a human to decide (product intent, deprecation candidates, performance targets)

Constraints
Do NOT modify any code during this audit. Analysis only.

Do not pad the report. If a dimension is healthy, say so in one sentence and move on.

Calibrate to the project's maturity. Don't recommend enterprise-grade infrastructure for a weekend prototype unless the owner's goals demand it.

Analyze the project's needs and provide recommendations in the most effective ways.

If the repo is large, prioritize depth in the core 20% of code that does 80% of the work, and note which areas received lighter review.

[ClaudeDevs @ClaudeDevs]

We've added an observability dashboard for developers of connectors.

Connectors let third-party developers bring their tools and data to Claude via MCP.

[@lowercasebryan @lower_case_b]

@helloparalegal What about data mining and guardrails?

[@lowercasebryan @lower_case_b]

@rileybrown I spend ~$4,000-5,000 monthly for extra usage alone. If/when Mythos is released, I'll absolutely do a $500-1,000 run to verify results.

[@lowercasebryan @lower_case_b]

@dani_avila7 Lots of false positives

[Claude @claudeai]

We've doubled usage limits in Claude Cowork for the next month.

Delegate bigger, more complex tasks to Claude.

[@lowercasebryan @lower_case_b]

@Jaxwilkerson29 @CoachVenables @OU_CoachB @OU_Football @JimmyWilkerson @Coach_Mike_D @CAHS_FOOTBALL20 @JRConrad64 @Josh_Scoop @ParkerThune @Rivals Got a little choked up reading this. He's absolutely proud and smiling down!

[Gabe Pereyra @gabepereyra]

Last week we held Harvey Hacks, our internal hackathon.

27 projects total across our 200-person eng team.

Wanted to highlight a few hackathon projects:

[@lowercasebryan @lower_case_b]

@agent_clerk Now this is a smart feature. Good job.

[Anthropic @AnthropicAI]

Our internal data shows Claude is accelerating AI development—a possible path to recursive self-improvement, or AI autonomously building a more capable successor.

It’s happening faster than we thought, and the implications deserve greater attention. https://t.co/OVVPJO7VQx

[@lowercasebryan @lower_case_b]

@StorminInNorman @Geraldini93 Would love to see Gerald enter coaching. I believe he'd be a recurring juggernaut, and his real-life experiences would help shape every young man on the team. Don't be scurred Gerald, come on home!

[Clerk @agent_clerk]

Clerk is low risk by design. Here is why our architecture is bulletproof:

1. Zero Copyright Risk

Under 17 U.S.C. § 105, federal court records are public domain. Opinions, dockets, and filings cannot be copyrighted. Building on Clerk carries the exact same legal risk as building a weather app on top of https://t.co/YyoZGEvJBg 👉 zero.

2. Bulletproof Compliance

Clerk delivers legal information, not legal advice. Every single card carries this explicit disclaimer. This is the same regulatory playbook Westlaw and Bloomberg Law have used for half a century, fully insulating Clerk from Unauthorized Practice of Law (UPL) liabilities.

3. Structural Sustainability

We don't rely on private gatekeepers or restrictive vendor licenses. Because the underlying data is public federal records, our infrastructure is provider-agnostic. No one can change the terms of service on us.

4. Scale is a Tailwinds, Not a Threat

Growth anxiety here is upside down. More AI agents querying public records isn't a strain on the system, it’s exactly how public data was always meant to flow.

https://t.co/AbHLeI16dd

[Harrison Chase @hwchase17]

Verifiers are important for scaling evals/RL

But costs add up! So can we make them cheaper?

Some great work by @Vtrivedy10 @jakebroekhuizen in conjunction with @nikogrupen @gabepereyra and the Harvey team on this