agentsclimarketplace

Issta related work

Skill brycewang-stanford/Awesome-Journal-Skills/ISSTA-Skills/skills/issta-related-work

Use when positioning an ISSTA submission against the testing and analysis literature, covering the sibling venues (ICSE, FSE, ASE, ICST, PLDI, CAV), delta-first framing against the nearest technique, double-anonymous self-citation, distinguishing the closest tool baseline, and avoiding the venue-confusion of citing a paper to the wrong conference.From its SKILL.md

Install
npx -y skills add brycewang-stanford/Awesome-Journal-Skills --skill issta-related-work

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

3.7 KB, 783 tokens by cl100k_base, as published. Nobody here has run it

ISSTA Related Work

Use this to audit novelty and positioning. ISSTA reviewers know the testing/analysis literature well, so the job is to name the nearest technique and state the delta, not to survey the field. Reopen the current call for any dual-submission and prior-publication rules before advising authors.

Positioning checks

  • Separate a genuine technique or measurement contribution from an engineering improvement: a new bug class found, a new property analyzed, a new evaluation of a technique class, or a sharper-but-cheaper analysis.
  • Name the single nearest technique and state the delta in one sentence: what it could not do, what yours does, and at what cost. Reviewers look for the closest competitor first.
  • Sweep the right venues. Testing and analysis work is spread across ISSTA and its siblings; missing a paper because it was at FSE or CAV rather than ISSTA reads as not knowing the field.
  • Keep self-citations double-anonymous: cite your own prior tool in the third person and do not link to an identity-revealing repository.
  • Declare any overlap with a prior workshop or preprint version, and do not put duplicate archival work under review.

Literature lanes for testing and analysis

LaneTypical venuesWhat ISSTA reviewers check
Software-engineering flagshipsICSE, ESEC/FSE, ASEWhether the closest SE technique is compared or distinguished
Testing-specific venuesISSTA, ICST, ISSREWhether the direct predecessor and its benchmark are used
Programming languages / verificationPLDI, POPL, OOPSLA, CAV, TACASWhether the analysis's formal neighbours are acknowledged
Security testingS&P, CCS, USENIX Security, NDSSWhether fuzzing/analysis-for-security work is credited where relevant

A bibliography that cites only ISSTA papers signals to a reviewer that nearer work at FSE, ASE, or CAV was missed — a recognizable weakness that benchmark strength does not repair.

Delta-first positioning vignette

Suppose the paper proposes a directed fuzzer for a specific bug class. Its nearest neighbours: a coverage-guided fuzzer at a security venue with no directedness, a directed symbolic-execution tool at a PL venue that does not scale, and a prior ISSTA fuzzer for a different bug class. The novelty sentence names all three deltas — directedness the coverage fuzzer lacked, scale the symbolic tool lacked, and a different, harder target than the prior ISSTA fuzzer — rather than a generic "we improve on prior fuzzers."

Venue-attribution discipline

  • Before citing a paper as ISSTA, confirm the venue on dblp or the ACM DL; many canonical testing/analysis papers (KLEE at OSDI, EvoSuite at ESEC/FSE, Randoop at ICSE) are not ISSTA, and mis-attributing them signals carelessness.
  • For a tool everyone knows by name, cite the paper that introduced it, at its actual venue, not the venue where you first encountered it.
  • When unsure whether a venue is archival for dual-submission purposes, declare the overlap in the submission form rather than guessing.

Output format

[Novelty] clear / incremental / at-risk
[Closest lanes] <SE / testing / PL-verification / security>
[Nearest 3 works] <work -> venue -> delta>
[Venue-attribution check] <any paper cited to the wrong conference?>
[Archival-overlap risk] none / issues
[Novelty sentence] <ISSTA-ready delta statement>

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.