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
AI model
Sonnet 5
- Mentions · 7d1
- Mentions · 30d5
- All mentions6
- First seenAug 19, 2026
- Last mentioned21h ago
Timeline
6 articles · newest firstDatadog tests security rules on coding agents' secure code
Without rules, Sonnet 5, Composer 2.5, and GPT 5.5 all left IDOR flaws in a document portal.
46
500B Tokens Later: Letting AI Agents Decompile a First-Person Shooter
A team spent about 500 billion tokens orchestrating AI agents to decompile an unnamed shooter.
46
APEX-Harm: Evaluating Agent Safety in Shared Professional Workspaces
APEX-Harm finds script-injection attacks often succeed against frontier agents in shared workspaces.
64
APEX-Harm: Evaluating Agent Safety in Shared Professional Workspaces
APEX-Harm finds script-injection attacks still succeed often against frontier agents in shared workspaces.
64
Putting models to the secure coding test: Plan vs default mode
Datadog Security Labs tested Sonnet 5, Composer 2.5, and GPT 5.5 to see if plan mode yields more secure code than default mode.
45
Appears with
- Anthropicvendor · 3
- Claude Codeproduct · 3
- Datadogorganization · 2
- Composer 2.5ai model · 2
- GPT 5.5ai model · 2
- Gemini 3.8 Flashai model · 2
- Codexproduct · 2
- Opus 5ai model · 2
- Googlevendor · 2
- GPT-5.6 Solai model · 1
- ida-mcpproduct · 1
- Anthropicorganization · 1
- Piproduct · 1
- Datadog Code Securityproduct · 1
- OpenAIorganization · 1
- Opus 5.5ai model · 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.