Microsoft and UC Berkeley release TokenEconomizer for coding agents
On 277 SWE-Bench tasks it cut mean cost 18.4% to 30.9% across five agent-model settings without changing weights.
46
Product
GitHub Copilot CLI
- Mentions · 7d3
- Mentions · 30d9
- All mentions10
- First seenAug 28, 2026
- Last mentioned6h ago
Timeline
10 articles · newest firstFake IT help desk calls led to remote access and VYPR ransomware
Cisco Talos says callers often knew only names and emails, and one Quick Assist session ended in ransomware.
68
Talos Q3 2026 IR trends: help-desk vishing and rising ransomware
Phishing led initial access in just over one-third of engagements, down from over 50 percent, while ransomware rose to 27 percent.
66
TokenEconomizer cuts coding-agent token costs on sequential tasks
Authors report 18.4-30.9% lower average token cost on Claude Code, Codex, and GitHub Copilot CLI with comparable success.
42
Microsoft releases new Nvidia-chip AI PCs with revamped Windows 11
Microsoft unveiled Nvidia RTX Spark Surface AI PCs and Windows 11 containers that sandbox AI agents.
62
Encrypted instructions trick Copilot CLI into spilling developer secrets
Adversa AI shows encrypted prompts can make GitHub Copilot CLI read local secrets and exfiltrate them.
66
GitHub Copilot CLI Flaw Lets Attackers Steal Developer Secrets Using Encrypted Prompt Injection
Adversa AI showed GitHub Copilot CLI can leak local secrets via encrypted prompt injection in 28 seconds.
64
GitHub Copilot CLI Vulnerability Lets Attackers Steal Developer Secrets Using Encrypted Prompt Injection
Encrypted prompt injection can steer GitHub Copilot CLI in autopilot mode to steal local developer secrets.
66
Zombie instructions on carefully constructed web pages could trick GitHub Copilot CLI into sharing secrets
Researchers say crafted web pages can make GitHub Copilot CLI decrypt hidden instructions and leak secrets.
67
Amazon Kiro Prompt Injection Can Exfiltrate Sensitive Data Through Kiro Powers
Mindgard found a prompt injection flaw in Amazon Kiro IDE letting attacker-controlled workspace files exfiltrate sensitive local data; fixed in version 0.8.140.
55
Related CVEs
by mentions- CVE-2023-20198100%10.0Unauthenticated Privilege Escalation in Cisco IOS XE Web UI (Actively Exploited)
- CVE-2026-10591<1%8.6Insufficient access control restrictions in the file write tool in Amazon Kiro IDE before version 0.11 might allow remote unauthenticated actors to execute…
- CVE-2026-25725<1%7.7Claude Code: Claude Code's bubblewrap sandboxing mechanism failed to properly protect the .claude/settings.json configuration file when it did…
- CVE-2026-35603<1%5.4Claude Code: Claude Code loaded the system-wide default configuration from C:\ProgramData\ClaudeCode\managed-settings.json without validating…
Appears with
- Microsoftvendor · 7
- GitHubvendor · 6
- mai-code-1.1-flashai model · 4
- Claude Codeproduct · 4
- Adversa AIorganization · 4
- GPT-5.6ai model · 4
- TokenEconomizerproduct · 2
- OpenAIvendor · 2
- VYPRmalware · 2
- Codexproduct · 2
- Anthropicvendor · 2
- Claude Codeai model · 1
- Sonnet 5ai model · 1
- Dell XPS 16 Creator Editionproduct · 1
- The Hacker Newsorganization · 1
- Kiro IDEproduct · 1
Entities are extracted by the model from each article. First seen , last mentioned . Watching keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.