Limits

Where review allowances may be worth increasing.

Saved stage runs only; live work and overwritten retry attempts are excluded. Reaching a cap does not prove that review quality suffered. Compare affected reports before changing allowances. Skipped stages are excluded.

2,285 historical stage runs without limit telemetry excluded.

Token and tool limits

Work limited counts observed output truncations, forced final answers or refused tool calls. The tool allowance also bounds model turns, so a forced final answer can occur below the tool call cap. Reached cap counts measured usage at or above the cap. Fractions use runs with known observations only. Output tokens and web searches show the maximum in one response, not the stage total. Web searches also consume the total tool allowance; reaching that smaller remaining allowance is counted under tool calls. Providers do not report whether the web search cap stopped additional searches, so work limited is unknown.

Profile	Stage	Model	Limit	Configured cap	Saved runs	Work limited / observed	Reached cap / measured	Maximum used
default	adversarial	gpt-6-astra	Discussion lookups	8	126	0 / 126	0 / 126	2
default	adversarial	gpt-6-astra	History lookups	8	126	0 / 126	0 / 126	0
default	adversarial	gpt-6-astra	Output tokens per response	25,000	126	0 / 126	0 / 126	3,606
default	adversarial	gpt-6-astra	Tool calls / turns	72	126	13 / 126	13 / 126	72
default	adversarial	gpt-6.1-sol	Discussion lookups	8	123	0 / 123	0 / 123	6
default	adversarial	gpt-6.1-sol	History lookups	8	123	0 / 123	0 / 123	0
default	adversarial	gpt-6.1-sol	Output tokens per response	25,000	123	0 / 123	0 / 122	7,687
default	adversarial	gpt-6.1-sol	Tool calls / turns	48	28	17 / 28	17 / 28	48
default	adversarial	gpt-6.1-sol	Tool calls / turns	72	95	10 / 95	10 / 95	72
default	adversarial_claude	claude-opus-5-5	Discussion lookups	8	223	0 / 223	0 / 223	5
default	adversarial_claude	claude-opus-5-5	History lookups	8	223	0 / 223	0 / 223	4
default	adversarial_claude	claude-opus-5-5	Output tokens per response	64,000	223	0 / 223	0 / 223	35,837
default	adversarial_claude	claude-opus-5-5	Tool calls / turns	48	2	1 / 2	1 / 2	48
default	adversarial_claude	claude-opus-5-5	Tool calls / turns	72	221	19 / 221	19 / 221	72
default	adversarial_gemini	google/gemini-4-argon	Discussion lookups	8	26	0 / 26	1 / 26	8
default	adversarial_gemini	google/gemini-4-argon	History lookups	8	26	0 / 26	0 / 26	5
default	adversarial_gemini	google/gemini-4-argon	Output tokens per response	32,000	26	0 / 26	0 / 26	9,360
default	adversarial_gemini	google/gemini-4-argon	Tool calls / turns	48	26	18 / 26	18 / 26	48
default	alternatives	gpt-6-luna	Output tokens per response	4,000	330	0 / 330	0 / 330	771
default	alternatives	gpt-6-luna	Tool calls / turns	12	330	1 / 330	1 / 330	12
default	archaeologist	gpt-6-luna	Discussion lookups	8	348	52 / 348	93 / 348	8
default	archaeologist	gpt-6-luna	History lookups	8	348	0 / 348	0 / 348	1
default	archaeologist	gpt-6-luna	Output tokens per response	4,000	348	0 / 348	0 / 342	1,773
default	archaeologist	gpt-6-luna	Tool calls / turns	12	348	193 / 348	201 / 348	13
default	archaeologist	gpt-6-luna	Web searches per response	2	348	Unknown	30 / 348	3
default	build	gpt-6-luna	Output tokens per response	4,000	142	0 / 142	0 / 140	842
default	build	gpt-6-luna	Tool calls / turns	24	34	0 / 34	0 / 34	15
default	build	gpt-6-luna	Tool calls / turns	48	108	0 / 108	0 / 108	26
default	challenger	gpt-6.1-sol	Discussion lookups	8	43	0 / 43	0 / 43	5
default	challenger	gpt-6.1-sol	History lookups	8	43	0 / 43	0 / 43	2
default	challenger	gpt-6.1-sol	Output tokens per response	25,000	43	0 / 43	0 / 42	2,334
default	challenger	gpt-6.1-sol	Tool calls / turns	12	4	4 / 4	4 / 4	12
default	challenger	gpt-6.1-sol	Tool calls / turns	24	6	5 / 6	5 / 6	24
default	challenger	gpt-6.1-sol	Tool calls / turns	48	33	10 / 33	10 / 33	48
default	collator	gpt-6-luna	Output tokens per response	6,000	224	0 / 224	0 / 224	5,063
default	concurrency	gpt-6-luna	Output tokens per response	8,000	210	0 / 210	0 / 207	7,327
default	concurrency	gpt-6-luna	Tool calls / turns	12	4	0 / 4	0 / 4	8
default	concurrency	gpt-6-luna	Tool calls / turns	24	21	0 / 21	0 / 21	10
default	concurrency	gpt-6-luna	Tool calls / turns	48	185	2 / 185	2 / 185	48
default	dead_code	gpt-6-luna	History lookups	8	340	0 / 340	0 / 340	0
default	dead_code	gpt-6-luna	Output tokens per response	8,000	340	0 / 340	0 / 338	4,176
default	dead_code	gpt-6-luna	Tool calls / turns	48	340	5 / 340	5 / 340	48
default	design	gpt-6-luna	Output tokens per response	25,000	309	1 / 309	1 / 306	25,000
default	design	gpt-6-luna	Tool calls / turns	12	18	14 / 18	14 / 18	12
default	design	gpt-6-luna	Tool calls / turns	24	46	31 / 46	31 / 46	24
default	design	gpt-6-luna	Tool calls / turns	48	245	41 / 245	41 / 245	48
default	feedback	gpt-6-luna	Output tokens per response	4,000	245	0 / 245	0 / 244	2,514
default	independent	gpt-6-luna	Discussion lookups	8	348	0 / 348	0 / 348	1
default	independent	gpt-6-luna	History lookups	8	348	0 / 348	0 / 348	0
default	independent	gpt-6-luna	Output tokens per response	4,000	348	0 / 348	0 / 346	998
default	independent	gpt-6-luna	Tool calls / turns	12	45	1 / 45	1 / 45	12
default	independent	gpt-6-luna	Tool calls / turns	24	56	0 / 56	0 / 56	13
default	independent	gpt-6-luna	Tool calls / turns	48	247	0 / 247	0 / 247	45
default	major_verifier	gpt-6.1-sol	History lookups	8	34	0 / 34	0 / 34	3
default	major_verifier	gpt-6.1-sol	Output tokens per response	25,000	34	0 / 34	0 / 34	3,809
default	major_verifier	gpt-6.1-sol	Tool calls / turns	48	8	0 / 8	0 / 8	25
default	major_verifier	gpt-6.1-sol	Tool calls / turns	72	26	0 / 26	0 / 26	34
default	public_contract	gpt-6.1-sol	Output tokens per response	25,000	258	0 / 258	0 / 256	1,726
default	public_contract	gpt-6.1-sol	Tool calls / turns	24	34	0 / 34	0 / 34	19
default	public_contract	gpt-6.1-sol	Tool calls / turns	48	224	0 / 224	0 / 224	32
default	public_contract	gpt-6.1-terra	Output tokens per response	25,000	1	0 / 1	0 / 1	167
default	public_contract	gpt-6.1-terra	Tool calls / turns	48	1	0 / 1	0 / 1	12
default	router	gpt-6-luna	Output tokens per response	4,000	91	0 / 91	0 / 90	693
default	router	gpt-6-luna	Tool calls / turns	0	10	0 / 10	10 / 10	0
default	router	gpt-6.1-sol	Output tokens per response	8,000	233	0 / 233	0 / 233	1,718
default	router	gpt-6.1-sol	Tool calls / turns	8	233	32 / 233	32 / 233	8
default	script	gpt-6.1-sol	History lookups	8	24	0 / 24	0 / 24	0
default	script	gpt-6.1-sol	Output tokens per response	25,000	24	0 / 24	0 / 24	3,876
default	script	gpt-6.1-sol	Tool calls / turns	48	24	5 / 24	5 / 24	48
default	state	gpt-6-luna	Output tokens per response	4,000	210	0 / 210	0 / 208	657
default	state	gpt-6-luna	Tool calls / turns	12	1	0 / 1	0 / 1	11
default	state	gpt-6-luna	Tool calls / turns	24	15	0 / 15	0 / 15	12
default	state	gpt-6-luna	Tool calls / turns	48	194	0 / 194	0 / 194	31
default	tests	gpt-6-luna	Output tokens per response	4,000	339	0 / 339	0 / 337	818
default	tests	gpt-6-luna	Tool calls / turns	12	36	0 / 36	0 / 36	11
default	tests	gpt-6-luna	Tool calls / turns	24	55	0 / 55	0 / 55	16
default	tests	gpt-6-luna	Tool calls / turns	48	248	0 / 248	0 / 248	22
default	verifier	claude-opus-5-5	Discussion lookups	8	258	2 / 258	3 / 258	8
default	verifier	claude-opus-5-5	History lookups	8	258	0 / 258	0 / 258	7
default	verifier	claude-opus-5-5	Output tokens per response	64,000	258	0 / 258	0 / 257	43,924
default	verifier	claude-opus-5-5	Tool calls / turns	72	258	3 / 258	3 / 258	72
default	verifier	claude-opus-5-5	Web searches per response	2	258	Unknown	10 / 258	6
default	verifier	gpt-6-luna	Discussion lookups	8	90	5 / 90	7 / 90	8
default	verifier	gpt-6-luna	History lookups	8	90	0 / 90	0 / 90	0
default	verifier	gpt-6-luna	Output tokens per response	128,000	90	0 / 90	0 / 84	43,408
default	verifier	gpt-6-luna	Tool calls / turns	48	37	17 / 37	17 / 37	48
default	verifier	gpt-6-luna	Tool calls / turns	72	53	8 / 53	9 / 53	74
default	verifier	gpt-6-luna	Web searches per response	2	90	Unknown	10 / 90	12
default	write_verifier	gpt-6.1-sol	History lookups	8	23	0 / 23	0 / 23	0
default	write_verifier	gpt-6.1-sol	Output tokens per response	12,000	23	0 / 23	0 / 23	2,462
default	write_verifier	gpt-6.1-sol	Tool calls / turns	48	23	1 / 23	1 / 23	48

Recent affected stage runs

Newest job IDs first. Reports show the recorded limits for each stage.

Review	Profile	Stage	Observation	Report
#36704 / job 1788	default	challenger	Tool calls / turns: work limited	report
#36437 / job 1787	default	archaeologist	Tool calls / turns: work limited, Discussion lookups: reached cap	report
#36703 / job 1786	default	challenger	Tool calls / turns: work limited	report
#36594 / job 1784	default	archaeologist	Tool calls / turns: reached cap, Web searches per response: reached cap	report
#36702 / job 1783	default	router	Tool calls / turns: work limited	report
#36702 / job 1783	default	archaeologist	Tool calls / turns: work limited	report
#36167 / job 1782	default	router	Tool calls / turns: work limited	report
#36167 / job 1782	default	archaeologist	Tool calls / turns: work limited, Discussion lookups: work limited	report
#36167 / job 1782	default	adversarial_claude	Tool calls / turns: work limited	report
#36701 / job 1781	default	archaeologist	Tool calls / turns: work limited	report
#36641 / job 1780	default	archaeologist	Tool calls / turns: work limited	report
#36618 / job 1778	default	archaeologist	Tool calls / turns: work limited	report
#36618 / job 1778	default	design	Tool calls / turns: work limited	report
#36618 / job 1778	default	challenger	Tool calls / turns: work limited	report
#29491 / job 1774	default	script	Tool calls / turns: work limited	report
#29491 / job 1774	default	archaeologist	Tool calls / turns: work limited, Discussion lookups: reached cap	report
#36699 / job 1773	default	archaeologist	Tool calls / turns: work limited	report
#36699 / job 1773	default	design	Tool calls / turns: work limited	report
#36699 / job 1773	default	challenger	Tool calls / turns: work limited	report
#35728 / job 1772	default	router	Tool calls / turns: work limited	report