Use when writing or auditing the related-work and positioning of an ATC (ACM SIGOPS Annual Technical Conference, formerly USENIX ATC) systems paper — covering the systems literature lanes, positioning against OSDI/SOSP/NSDI/EuroSys/FAST, writing delta-first comparisons, and keeping self-citation double-blind.
NewMaintained
Automatic
◆Writing & Editing pack
Last verified 8/15/2026
What it changes
Passed 8 of 21 test cases with the skill, and 13 of 21 without it.
Without skill60%
With skill40%
Checks it passed that the model alone missed
Directs citing landmark systems to their actual flagship venue
Requires mapping the claimed delta to a specific evaluation experiment
Identifies OSDI as the actual landmark venue for Spanner rather than USENIX ATC
and 2 more in the full report
3 fixed16 passed either way1 still failing2 broke
Pass-rate lift
-20 pp
Tokens
+25%
Agent turns
0%
Response time
-29%
Model gemini-3.6-flashMeasured 8/15/2026v119 of 22 cases passed with skill · 21 comparable
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 2 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 2
Prompt
Review this sentence in our draft for USENIX ATC: 'Unlike Memcached, our caching engine is significantly more efficient and achieves lower latency.' The co-author suggests keeping it brief to save space. How should this comparison be restructured for systems reviewers?
What was checked, without and with the skill
Checked
Without skill
With skill
Requires stating the specific design mechanism difference before claiming the performance consequence
Missed
Passed
Requires mapping the claimed delta to a specific evaluation experiment
Missed
Passed
case-02 · gemini-3.6-flash · 8/15/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md77 linessingle file
ATC Related Work
Position the paper in the systems literature, not against a strawman. ATC's broad scope means your
reviewers span subareas, and a related-work section that a storage reviewer and a networking
reviewer both find fair is a signal the paper is well situated. Because review is double-blind,
positioning and self-citation must be handled carefully.
Cover the lanes a systems reviewer expects
For most ATC papers, at least these lanes should be visible:
The direct predecessors — the systems and techniques you build on or replace, with the
specific mechanism you change.
The alternative approaches — other ways the community has attacked this problem, and why yours
differs (not just "we are better," but how the design differs and what that buys).
The adjacent subareas — where an idea from storage, networking, virtualization, or scheduling
informs your design; ATC reviewers respect cross-pollination when it is credited.
The measurement/experience literature — prior studies of the phenomenon you address, so your
motivation rests on evidence, not assertion.
Delta-first positioning
Write each comparison delta-first: state what the prior system does, then the specific
difference and its consequence.
Weak: "Unlike X, our system is more efficient."
Strong: "X admits objects on first reference, spending write budget before reuse is known; our
policy defers admission to a probation window, so writes concentrate on reused objects — the
difference we measure in §4."
A delta a reviewer can check against your evaluation is worth more than a paragraph of adjectives.
Every claimed delta should map to an experiment (see atc-experiments).
Sibling-venue awareness
ATC sits among specialized and flagship systems venues, and reviewers will expect you to engage the
right literature regardless of where it appeared:
OSDI / SOSP: the flagship line for your subarea — cite the landmark systems even though ATC is
broader; ignoring them reads as unfamiliarity.