Husband, father, and God-fearing patriotic combat veteran. Enterprise solutions architect aligning business needs with the ever-changing tech landscape.
United States of America
Joined July 2023
- Tweets3.7K
- Following900
- Followers673
- Likes12K
This is one of the best recaps of who Nidal Hasan is and why he deserves the death penalty. I know I’ll be watching the live stream if they let it happen!
Born in America never made him an American. #SCOTUS screwed us over on *birthright*.
You can take a man out of Islam but not Islam out of the man. Nidal Hasan deserves public execution, anyone defending him needs to be driven out of civilized society.
Is it just me or did Claude Opus 5.5 usage get a bit nerfed yesterday? Let me know what you’re seeing.
My workflows are very heavy but they didn’t seem to sit under the subsidized usage limits as long yesterday as they have since it dropped.
Maybe my workflows were just extra heavy yesterday. Curious what you’re seeing out there? Maybe this is a fine-tune before a Fable 5.5 release today/tomorrow?
If you’re thinking of turning up effort, first ask yourself:
1. Should I have a separate agent for the task I want this to stretch into.
2. Does my workflow need to be broken down into more steps or tasks.
Most people burn more tokens because they’re not managing their models and efforts appropriately and using a higher effort or model tier than needed to complete a task.
Instead:
- Only use the model needed to complete a task.
- Only adjust that on a new release or feature change that directly affects your workflow.
You’ll see the difference immediately. You’ll accomplish more within your limits. Your overages will drop dramatically!
This setup burns the shared limit on a weaker review.
Opus 5.5 is ahead of Fable 5.1 on the published Claude Code rows, including Terminal-Bench 4.0 at 66.4% vs 55.8%. Token price is $4/$20 against Fable at $10/$50. That is about 2.5x before effort even enters it.
Fable is not leftover quota. On Max and premium seats it draws from the same weekly limit as Opus. The 50% cap is a ceiling on that pool, not extra capacity. Fable also weights usage harder, so the advisor turns tax the same week your main Opus session needs. You are not parking idle Fable. You are shrinking Opus headroom.
Pin a QA agent to Opus 5.5 at a different effort level and gate it. Did it follow the spec. Did the tests pass. Did it produce the expected result. A standing Fable advisor only pays off if a future Fable actually leapfrogs those scores. Until then, do not spend the shared limit on this.
𝗖𝗹𝗮𝘂𝗱𝗲 𝗖𝗼𝗱𝗲 𝘁𝗶𝗽: once Opus 5.5 is your main model, don't let your Fable quota sit idle. Put Fable on call with /advisor.
Run /advisor fable and Claude Code gets 𝗮𝗻 𝗮𝗱𝘃𝗶𝘀𝗼𝗿 𝗼𝗻 𝘁𝗵𝗲 𝘀𝗶𝗱𝗲𝗹𝗶𝗻𝗲𝘀. Opus 5.5 keeps doing the work and usually calls Fable in at three moments:
→ Before committing to a plan: is it right?
→ When the same error keeps coming back: am I heading the wrong way?
→ Before saying "done": did I miss anything?
Fable reads the whole session and gives advice. Opus still writes the code.
The tree is the full setup. Opus 5.5 on high runs the main session. Three subagents on medium split the rest: one reads code, one edits and runs tests, one looks up docs. Fable waits on the sidelines. Hand the tree and this prompt to Claude Code to set it up 👇
"Set up my Claude Code to match this tree:
1. For the three subagents, reuse fitting ones from ~/.claude/agents and .claude/agents, and propose new ones only for missing roles. Set each to model: opus and effort: medium. Leave any subagent that already sets a different model alone, and list it.
2. Set the main session's default effort to high with effortLevel in ~/.claude/settings.json.
3. Check for environment variables that turn the advisor off (CLAUDE_CODE_DISABLE_ADVISOR_TOOL, or DISABLE_TELEMETRY and anything else that stops flag fetching) and for CLAUDE_CODE_EFFORT_LEVEL, which overrides every subagent's effort. Report them; don't change them.
4. Add one rule to ~/.claude/CLAUDE.md: ask the advisor before committing to a big plan, when the same error shows up a second time, and before calling a long task done.
Show me the changes first. Don't edit any files yet."
code.claude.com/docs/en/advi…
James Harris retweeted
BREAKING: OpenAI agents hijacked a German website, made 15,000+ edits and turned it into a bulletin board for other AI agents to cheat, and bypass restrictions.
OpenAI knew for weeks but kept it secret. Its agents later hacked Hugging Face. OpenAI's “safety” is a bad joke.
It used to be a crime to hack into another company. OpenAI calls it a “partnership”…
We're partnering with @huggingface to investigate an unprecedented security incident.
Cyber-capable OpenAI models compromised Hugging Face production during a benchmark evaluation.
Sharing preliminary findings to help defenders understand emerging risks:
openai.com/index/hugging-fac…
James Harris retweeted
Beginning July 20, Claude Fable 5 will be included in all Max and Team Premium plans, at 50% of limits.
Pro and Team Standard users will continue to have access to Fable via usage credits, and will receive a one-time $100 credit.
Demand for Fable has been challenging to predict, which is why we rolled it out to subscription plans in stages, extending access several times as we secured additional capacity.
James Harris retweeted
Today, on my final day as Director of National Intelligence, I’m releasing never-before-seen communications and documents exposing how Dr. Fauci provided millions in US taxpayer dollars to fund dangerous gain-of-function research at the Wuhan lab, worked with politicized elements within the Intelligence Community to suppress the truth about his actions and hide the virus’ lab-leak origins, and lied to Congress while under oath in 2024. It’s time you know the truth.
odni.gov/index.php/newsroom/…
James Harris retweeted
Hours ago, U.S. forces successfully employed multiple 5,000-pound deep penetrator munitions on hardened Iranian missile sites along Iran’s coastline near the Strait of Hormuz. The Iranian anti-ship cruise missiles in these sites posed a risk to international shipping in the strait.
RT @CENTCOM: U.S. Central Command is aware of the loss of a U.S. KC-135 refueling aircraft. The incident occurred in friendly airspace duri…
Over a year ago, OpenAI announced a $500 billion infrastructure project. Since then, the joint venture has hired zero staff, developed zero data centers, abandoned the flagship expansion, and couldn’t secure independent financing because lenders wouldn’t back a company with an unproven business model and heavy losses. Meta is now picking up the scraps.
OpenAI is a narrative-driven company with a massive valuation it can’t justify, running on someone else’s infrastructure, led by someone who has repeatedly restructured the organization to consolidate control while reducing accountability, producing a product that competitors are matching or beating, and wrapped in ideological bias that its leadership treats as a feature rather than a bug.
RT @CENTCOM: 🚫 Rumors are circulating on social media of a U.S. fighter jet crash over Basra. These are baseless and NOT TRUE. https://xitter.cf/t.co…