Tools for a one-off TypeScript and TSX code rewrite: ESLint, Semgrep, ast-grep and codeedit

Choose a custom ESLint rule for a reviewed TypeScript/TSX rewrite, by a narrow margin. Compare Semgrep, ast-grep and codeedit by matching precision, failure handling and setup effort.

PRACTICAL PICKSPreferredPickOur Preferred PickReviewed custom TS/TSX rewrite$0

ESLint

Use a small custom rule with the free TypeScript parser. Preview with --fix-dry-run --format json --exit-on-fatal-error before applying. A narrow lead: 4.05 against Semgrep 4.03 and ast-grep/codeedit 4.00.

Needs plugin and parser setup; no interactive approval or whole-batch rollback on apply.

Why ESLint is our Pick
Newcomer assessment

No Rising Star selected

Our November 7 search found usable rewriting tools, but did not establish every required launch-date, platform, task and safety fact for a newcomer award.

An unknown first-release date blocks the award, not comparison coverage.

Who is this article for? Developers changing one deprecated JSX prop and direct call shape in a small, versioned TypeScript/TSX repository.

About 19 min read

What Preferred Pick, Rising Star and Ready to Rock mean

The quick list

Start with a custom ESLint rule and dry parsing-error gate, our narrow Pick. Consider Semgrep for existing pattern-rule workflows, ast-grep for short patterns and interactive review, or codeedit for syntax-checked edits.

The compared local routes are free, checked November 7, 2026: ESLint with its MIT TypeScript parser, Semgrep Community, ast-grep and codeedit. Setup and reviewing results still take developer time.

The comparison at a glance

Compare how each tool reads TypeScript and TSX, how you review a change and recover from a bad run, and how actively each is maintained. Select up to three tools to compare.

ESLint leads narrowly for this one-off rewrite, followed by Semgrep. ast-grep and codeedit tie; their display order does not imply a lead.

Local tools for a one-off TypeScript and TSX rewrite
FeatureESLintSemgrepast-grepcodeedit
Best fitOur Pick: custom rule with a dry parsing-error gateExisting pattern-rule workflows; preview scan before applyShort patterns and interactive diff review in a repository you trustSyntax-checked edits; tied with ast-grep
Review before writing--fix-dry-run --format json returns fixed source without writingscan --autofix --dryrun prints changes without writingDiff by default; --interactive asks before each change-n previews the diff; Python rewrite dry_run available
Files that do not parse--exit-on-fatal-error exits 2; use with dry run before applyscan --strict exits 3 for invalid syntax; use with dryrunParser recovers and may still rewrite; no stop optionMaker says edits adding syntax errors are rejected; no type check
Latest release10.14.0, October 30, 20261.184.0, November 5, 20260.51.1, November 2, 20261.2.1, October 28, 2026
Main limitationPlugin/parser setup and no whole-batch rollback on applyExperimental autofix with explicit data-loss warningNo type or scope checks; can edit files that do not parseNo type/import resolution; maker write guarantees not independently tested

ESLint is our Pick for this reviewed one-off rewrite, narrowly: 4.05 against Semgrep 4.03 (4.025 before rounding) and ast-grep/codeedit 4.00 each. ast-grep and codeedit remain tied; display order is not a lead. Scores express documented task fit, not measured speed or accuracy. Preview on the exact paths, keep a separate copy of current files, and review the final diff; a dry parsing-error gate does not make apply atomic. codeedit write guarantees are maker claims.

All features and sources 14 criteria

Open the groups that matter to you. They compare the same selected tools. “Planned” means future work, not an available feature. Sources include the published conditions.

Matching and effort

Show 5 criteria
Matching and effort
FeatureESLintSemgrepast-grepcodeedit
Best fitOur Pick: custom rule with a dry parsing-error gateExisting pattern-rule workflows; preview scan before applyShort patterns and interactive diff review in a repository you trustSyntax-checked edits; tied with ast-grep
Price and licence$0; MIT ESLint and TypeScript parser; no account$0 local Community CLI and own rule; no account$0; MIT; no account$0; MIT source and bundled grammars; no account
How it reads TypeScript and TSXAST visitors and exact-range fixers with typescript-eslintTSX parser with syntax patterns and rule fixesSeparate TypeScript and TSX syntax-tree grammarsSeparate TypeScript and TSX Tree-sitter grammars and byte edits
Knows which import a name refers toPer-file scope API; custom rule must check itNo; the local Community route matches names without resolving their TypeScript importNo; its FAQ says no type or scope analysisNo documented type or import resolution
What you writeCustom plugin, fixer and TypeScript parser configurationLocal YAML pattern/fix ruleA one-line pattern or short YAML ruleA short syntax pattern and replacement; raw queries if needed
Sources for matching and effort

ESLint custom rules, CLI and parser licence; Semgrep scan flags and exit codes and releases; codeedit README and releases, checked November 7, 2026. ast-grep rewrite guide and FAQ, checked October 19; release fixes, checked November 7. Maker behaviour claims are not independently tested. Earlier-check evidence describes October 1 only.

Review and failure handling

Show 4 criteria
Review and failure handling
FeatureESLintSemgrepast-grepcodeedit
Review before writing--fix-dry-run --format json returns fixed source without writingscan --autofix --dryrun prints changes without writingDiff by default; --interactive asks before each change-n previews the diff; Python rewrite dry_run available
Files that do not parse--exit-on-fatal-error exits 2; use with dry run before applyscan --strict exits 3 for invalid syntax; use with dryrunParser recovers and may still rewrite; no stop optionMaker says edits adding syntax errors are rejected; no type check
How files are writtenOverlapping fixes compete; up to ten rounds; no whole-batch rollbackExperimental autofix; maker warns of data loss; no safe-write guaranteeTemporary file then replace since 0.51.0, October 15, 2026Maker says stale/overlapping edits rejected and multi-file writes all or nothing
Earlier hands-on check, October 1, 2026 (before our research-only method)Not checkedNot checked0.45.3: four expected edits on named files; broad scan also edited malformed fileNot checked
Sources for review and failure handling

ESLint custom rules, CLI and parser licence; Semgrep scan flags and exit codes and releases; codeedit README and releases, checked November 7, 2026. ast-grep rewrite guide and FAQ, checked October 19; release fixes, checked November 7. Maker behaviour claims are not independently tested. Earlier-check evidence describes October 1 only.

Maintenance and installation

Show 5 criteria
Maintenance and installation
FeatureESLintSemgrepast-grepcodeedit
Latest release10.14.0, October 30, 20261.184.0, November 5, 20260.51.1, November 2, 20261.2.1, October 28, 2026
Maintenance signalNamed steering committee/committers; regular releases and fixesActive team and release history; Community support best effortFrequent releases; rule-key fix shipped November 2Recent matcher and Windows build fixes; one principal contributor
Security recordOctober 30 dependency/fixer fixes; published security policyLatest-only security policy; dependency coverage not completeCritical/high configuration advisories fixed in 0.50.0Current advisory coverage unknown; no clean-history claim
InstallationNode.js, ESLint, TypeScript parser and custom configurationPython; Windows/WSL adds setup effortnpm, Homebrew, pip, cargo, Scoop or MacPortsNative Windows/macOS/Linux binaries or Python package
Main limitationPlugin/parser setup and no whole-batch rollback on applyExperimental autofix with explicit data-loss warningNo type or scope checks; can edit files that do not parseNo type/import resolution; maker write guarantees not independently tested
Sources for maintenance and installation

ESLint custom rules, CLI and parser licence; Semgrep scan flags and exit codes and releases; codeedit README and releases, checked November 7, 2026. ast-grep rewrite guide and FAQ, checked October 19; release fixes, checked November 7. Maker behaviour claims are not independently tested. Earlier-check evidence describes October 1 only.

What matters when choosing

ESLint is our Preferred Pick for a one-off structural rewrite in a small, versioned TypeScript/TSX repository, by a narrow margin. Write a small custom rule with exact-range fixes and the free typescript-eslint parser. The rule can target a Card prop and a renderLegacy(...) call without changing the same words in strings or comments; its scope API can check which binding a call uses within a file. Preview with --fix-dry-run --format json --exit-on-fatal-error: no files are written, JSON includes fixed source, and a fatal parsing error returns exit code 2. Inspect that output before running --fix on the same paths. Custom rules, CLI reference, typescript-eslint source and MIT licence, checked November 7, 2026.

The lead is narrow: ESLint scored 4.05 of 5, Semgrep 4.03, and ast-grep and codeedit 4.00 each. The unrounded gaps are 0.025 and 0.050, smaller than a one-point change on our heaviest criterion, which carries 30% of the weight. ESLint earns its lead from its maintenance record and documented dry parsing-error gate; it needs more rule and parser setup than ast-grep. There is no interactive approval, and the later apply is not a transaction: if one file fails, do not assume changes to other files are undone. Overlapping fixes compete, and fixes can run for up to ten rounds. Fixer guidance, releases, checked November 7, 2026.

Consider ast-grep if a short pattern and interactive diff review matter more than a parsing-error gate. Use version 0.51.1 or later in a repository you trust. It does no type or scope checks, so a matching name may refer to the wrong function, and its error-tolerant parser can edit files that do not parse. It shows a diff by default and writes through temporary files, but those protections do not catch wrong-binding matches. FAQ, rewrite guide, checked October 19; releases, checked November 7, 2026.

Semgrep fits teams already writing pattern rules; codeedit fits developers who want syntax-checked edits with a native CLI. Semgrep calls autofix experimental and warns of data loss; keep its dry run separate from apply. codeedit’s maker describes preview, rejection of syntax-breaking edits and all-or-nothing multi-file writes; these are documented claims, not our own test results. Neither tool resolves TypeScript types for you. Semgrep CLI, codeedit, checked November 7, 2026.

No Rising Star selected. Our November 7 search did not establish a newcomer meeting every launch-date, platform, task and safety requirement. An unknown first-release date prevents the award, not comparison coverage.

Whichever tool you use, copy current target files somewhere separate before applying, including uncommitted edits, review the full diff afterwards, and run your project’s own type check. A dry-run gate does not make a later apply atomic.

Disclosure: Practical Picks is researched and written with Claude, an AI model made by Anthropic. ast-grep documents Claude Code integrations; codeedit provides a Claude skill; Codemod documents a Claude harness. These integrations earned no credit, and we have no relationship with Anthropic or these makers. ast-grep integrations, checked October 19; codeedit, Codemod CLI, checked November 7, 2026.

ESLint

Preferred Pick · reviewed custom rewrite

$0MIT ESLint and typescript-eslint parser; local, no account

Use a small custom ESLint rule with a dry parsing-error gate before apply. It leads narrowly, but takes plugin and parser setup.

Reasons to choose it

  • Exact source-range fixers and per-file scope API
  • Dry run returns changed source without writing
  • Named steering committee, regular releases and security policy

What to watch

  • Needs a custom rule and TypeScript parser configuration
  • No interactive per-change approval
  • Apply is not a transaction; overlapping fixes compete

ESLint can express this rewrite as a local rule with a fixer. Its API visits syntax nodes, replaces specific source ranges and exposes variable scopes, so the rule can distinguish a call from a string or comment and inspect its binding within a file. TypeScript and TSX need the typescript-eslint parser; ESLint alone is not the complete installation. This is a credible option for a team that wants to keep enforcing the rule after the migration, but it takes a plugin and configuration rather than ast-grep’s short pattern. Custom rules, typescript-eslint, checked November 7, 2026.

Keep the fixer small and disable unrelated rules for the migration. ESLint’s documentation warns that overlapping fixes compete and that fixes may run for up to ten rounds. Use --fix-dry-run --format json --exit-on-fatal-error first: it returns changed source without writing files and exits 2 for a fatal parsing error. Inspect the output before --fix on the same paths. This gate does not provide whole-batch rollback; keep a separate copy of current files. ESLint and its required typescript-eslint parser are MIT-licensed and free. CLI reference, parser source, checked November 7, 2026. Fixer guidance, checked November 7, 2026.

Read the full ESLint analysis and sources

ESLint itself is MIT-licensed and has a named technical steering committee and several committers. Version 10.14.0, released October 30, 2026, includes fixes to both its fixers and a vulnerable dependency. That is evidence of current maintenance, not a guarantee that an arbitrary custom rule is safe. Repository, releases, security policy, checked November 7, 2026.

Used ESLint? Share your experience to help other readers choose.

Semgrep

Evaluated · existing pattern-rule workflows

$0Local Community CLI and own rule; no account

Consider Semgrep if your team already writes its pattern rules. Preview with scan before applying; the maker warns autofix is experimental.

Reasons to choose it

  • Syntax patterns and fixes support TSX
  • scan --autofix --dryrun previews without writing
  • --strict gives a nonzero exit on invalid syntax

What to watch

  • Autofix is experimental and carries a data-loss warning
  • No documented safe-write or whole-batch rollback guarantee
  • Python and Windows/WSL setup add effort

Semgrep’s Community CLI can run your own local rule without an account. Its TSX parser matches a JSX prop or direct call by syntax rather than text; put the replacement in the rule’s fix field. On this local Community route, matching a name does not establish which TypeScript import it refers to. Source, pricing, CLI reference, checked November 7, 2026.

Use semgrep scan with your local rule and --autofix --dryrun --strict first. The scan reference says dryrun prints proposed changes without writing, and strict returns exit code 3 for invalid source syntax. Review that output and resolve parsing errors before applying to the same paths. Do not substitute semgrep ci: its autofix and dryrun flags are documented as ignored. The maker calls scan autofix experimental and warns that data loss can occur, so keep a separate copy of current files. No safe-write or whole-batch rollback guarantee is established here. CLI reference, scan options and exit codes, checked November 7, 2026.

Read the full Semgrep analysis and sources

Active releases and a security policy support maintenance, but the local Community route has best-effort support. Python and Windows/WSL installation add effort compared with ast-grep’s native packages. Choose it when your team already works with Semgrep rules; for this isolated migration, ESLint’s custom rule is our narrow overall choice. Releases, security policy, README, checked November 7, 2026.

Used Semgrep? Share your experience to help other readers choose.

ast-grep

Also consider · short patterns and interactive review

$0Free, open source (MIT) · npm, Homebrew, pip, cargo

Choose ast-grep 0.51.1 or later for short patterns and interactive diff review in a repository you trust. It lacks type or scope checks and can edit files that do not parse.

Reasons to choose it

  • Shows a diff by default and can ask before each change
  • Separate TypeScript and TSX grammars skip strings and comments
  • Frequent releases; rule-key fix shipped November 2

What to watch

  • No type or scope checks
  • Can rewrite inside a file that does not parse
  • Critical configuration advisory fixed in 0.50.0; use trusted repositories

ast-grep is a command-line tool for structural search, linting and rewriting. Patterns look like ordinary code with $NAME wildcards, so a call rename can be one line. For a JSX attribute you write a short YAML rule that targets jsx_attribute nodes, as in the attribute-renaming example in its TSX catalog, and limit it to Card elements. TypeScript and TSX need distinct parsers in ast-grep; its languageGlobs setting can parse .ts files as TSX so that one rule covers both. Without --update-all it prints a diff and writes nothing, and --interactive asks before each change. Rewriting code, scan options, TSX catalog

Writing is safer than it was a month ago. Since 0.51.0, --update-all writes to a temporary file in the same folder and replaces the original only once the write completes. That closed a September 22, 2026 report that interrupting a run between opening and writing a file left it at 0 bytes. One gap remains. ast-grep’s FAQ says its parser “can recover from syntax errors and continue parsing the rest of the code”, and the scan command has no option to stop on files that fail to parse, so a rule can still change a matching part of a broken file. A separate September 24 report found that a misspelt rule key, such as ignore for ignores, was silently ignored, dropping a scope limit; a fix that rejects unknown keys was merged on October 19, 2026 and shipped in 0.51.1 on November 2. Release notes, checked November 7, 2026. Write fix, 0-byte report, FAQ, unknown-key report

Read the full ast-grep analysis and sources

It releases frequently; the latest release we checked was 0.51.1 on November 2, 2026. Releases, checked November 7, 2026. Its security policy provides fixes for the latest stable release and asks for private reports. On October 13, 2026 the maintainer published two advisories. The critical one (CVSS 9.6) covers versions before 0.50.0: customLanguages.libraryPath in a project’s sgconfig.yml could load a native library from outside the project. The high-severity one describes an untrusted configuration running native code when you start scan, run or the language server. Since 0.50.0, loading custom-language libraries from a project’s configuration needs explicit permission. This matters most if you run ast-grep on repositories you did not write, or if your editor starts its language server automatically. Security policy and advisories, critical advisory, high-severity advisory, releases

In an earlier hands-on check on October 1, 2026, before we adopted our research-only method, ast-grep 0.45.3 made exactly the four expected edits in two named files and left strings, comments and lookalike names alone. A broader scan of the whole source folder also changed a matching attribute inside a deliberately malformed file and still reported success. That version predates both the write fix and the advisory fixes. Earlier check record, October 1, 2026

Used ast-grep? Share your experience to help other readers choose.

codeedit

Evaluated · syntax-checked structural edits

$0MIT native CLI and bundled grammars; local, no account

Consider codeedit for short syntax patterns and validated edits. It ties ast-grep; its write guarantees are maker documentation, not our own test results.

Reasons to choose it

  • TypeScript and TSX patterns with unified diffs
  • Preview and rejection of stale or syntax-breaking edits
  • Prebuilt Windows, macOS and Linux binaries

What to watch

  • Syntax matching does not resolve types or imports
  • Maintenance concentrated on one principal contributor
  • First public release date and advisory coverage unknown

codeedit is a local Rust CLI and Python package built on Tree-sitter, with separate TypeScript and TSX support. It matches call or JSX syntax patterns, splices precise source edits and returns a unified diff. Preview before applying: the README shows -n for the CLI and dry_run=True for Python rewrites. Prebuilt Windows, macOS and Linux binaries avoid Python setup if you choose the native CLI. Its source and bundled grammar licences are MIT. Repository and installation, checked November 7, 2026.

The maker says it rejects stale, overlapping and syntax-breaking edits and writes multi-file changes all or nothing. We have not verified that guarantee in practice. Syntax matching does not establish type correctness or which import a name refers to. Inspect matched locations and run your TypeScript checks afterwards. README, checked November 7, 2026.

Read the full codeedit analysis and sources

The latest displayed release is 1.2.1 from October 28, 2026. Recent human commits address matcher backtracking and a Windows grammar build failure; maintenance is concentrated on one principal contributor. The earliest public release date is unknown in our evidence, so it receives no newcomer award. It ties ast-grep at 4.00: choose between documented syntax checks and interactive review rather than treating display order as a lead. Repository, releases and recent changes, checked November 7, 2026.

Used codeedit? Share your experience to help other readers choose.

Which product are we missing?

Know a product that belongs in this article? Tell us what it does differently and why it deserves a closer look. Newcomers welcome.

Readers and product teams can suggest candidates. Suggestions inform our research; they do not guarantee inclusion or a recommendation.

Suggest a product

Opens an email draft with this article attached as a link. Nothing is sent until you send it.

No email app? Send the product name, website, and this article’s title to admin@practicalpicks.guide.

What people are saying

We found no representative satisfaction data for any of these tools. Issue trackers show individual experiences, not failure rates. ast-grep’s two September 2026 reports, of a file left empty after an interrupted write and of a misspelt rule key silently ignored, were both closed by fixes within four weeks, with no written maintainer reply on either issue. 0-byte report, unknown-key report

jscodeshift’s 2024 maintenance announcement drew questions from other contributors about why one company was made a partner and whether the project needed a steering committee; the maintainer replied that “anyone is welcome to contribute” and that the project was too small for a committee. Comby users have been waiting since September 2025 for a build that works on current Debian, with “I run into the same issue now” added in December 2025. We did not find user reports about ts-morph that bear on this job. jscodeshift announcement, Comby Debian report

Download counts show footprint, not direct use. In the week of October 12 to 18, 2026, npm recorded about 17.9 million downloads of ts-morph, 3.4 million of jscodeshift and 1.1 million of ast-grep’s command-line package; many come from other tools that depend on them. npm download counts

How we researched

This is public-source research. For this review we did not install, run or test any tool, and no maker supplied access or payment. We read each project’s documentation, readme, release and commit history, issue tracker and security page, plus maintainer announcements, the OSV vulnerability database, Homebrew’s package data and npm download counts, and rechecked every claim on October 19, 2026. On November 7 we checked licences, ast-grep release fixes and additional candidate sources; this was not a complete refresh of every product fact.

Earlier, on October 1, 2026, before we adopted our research-only method, we ran one small local check on Linux: a synthetic set of TypeScript and TSX files with four expected edits, near misses in strings, comments and lookalike names, and a deliberately malformed file. jscodeshift 17.4.0, Comby 1.8.1 and ast-grep 0.45.3 each made exactly the four expected edits on named files; the failure probes are described in the linked record. We keep that record as dated evidence of how those versions behaved on that day. It is consistent with the documentation but did not decide the recommendation, and no new checks were run for this review. Earlier check record, October 1, 2026, earlier check fixture, transform and expected patch

We wrote this for a developer who needs to rename one deprecated JSX prop on one component and one function call across a small TypeScript/TSX repository kept in version control, while leaving strings, comments and lookalike names alone, with a preview and a way back from a bad run. Every recommended tool must be free, open source and usable locally without an account. Before scoring, we fixed four criteria and their weights: how precisely it rewrites TypeScript and TSX (30%); review and failure safety, including what happens to files that do not parse and how files are written (25%); maintenance and continuity (25%); and setup and effort for a one-off job (20%). ESLint scored 4.05 after its CLI documentation established a dry-run parsing-error gate; Semgrep scored 4.03 (4.025 before rounding), with ast-grep and codeedit tied at 4.00. The recommendation is narrow. Cost is information rather than a scoring criterion; all four compared local routes are free.

On November 7, 2026, we examined GitHub’s codemod topic and structural-search topic, the October 19 discovery record and available candidate repositories. No newcomer established all four eligibility conditions. zmod has releases and a TSX route, correcting our earlier no-release finding, but Windows/macOS support and first-release provenance are unknown. Codemod/JSSG, codeedit, nodeseek and Threadmoth have useful capabilities but no established earliest public release date in our evidence; nodeseek’s Windows support is also unknown. GritQL predates the window and its actual CLI is still alpha. codemod-js had a release in 2019. eslint-plugin-command and Hypermod Community did not establish the exact custom TSX route or first-release provenance. We could not identify a product behind the December 2025 example-based rewrite-engine announcement. This supports no award, not a claim that no newcomer exists. zmod, Codemod, nodeseek, Threadmoth, codemod-js, eslint-plugin-command, Hypermod Community, checked November 7, 2026.

Other fitting tools received the same criteria. Threadmoth scored 3.90: its ast-grep attachment adds guarded proposals and journalled recovery, with another binary and workflow to configure. Codemod scored 3.80 for semantic lookup and local dry run; neither a batch parse gate nor safe replacement is documented in the current CLI/JSSG pages. jscodeshift with Recast scored 3.78 and remains worth considering when a file that does not parse must stop the batch: --dry --print --fail-on-error previews transformed files and exits 1 on errors, but apply can leave other files changed. Recast preserves unchanged code; transform effort and stewardship uncertainty add costs. ts-morph scored 3.58 and suits compiler-level binding checks, with a user-built script and preview. nodeseek scored 3.55: syntax patterns and diff preview, but source-build setup and error-tolerant matching. Babel and zmod each display 3.53; zmod wins their equal-total tie on precision, while Babel needs more printer/script work. Putout scored 3.40 with concise plugins but an opinionated printer. Comby scored 2.43: compact templates do not overcome Homebrew’s January 13, 2027 disable date and the unresolved Debian 13 build failure. Codemod CLI, JSSG, Babel, Putout, checked November 7, 2026; jscodeshift, maintenance announcement, ts-morph, Comby packaging, Debian 13 build report, checked October 19, 2026.

codemod-js, eslint-plugin-command and Hypermod Community remain unscored: we could not confirm that they can apply this custom TSX prop and call change or preview it before writing. TypeScript editor rename refactors cannot migrate external declarations. OpenRewrite’s JavaScript route needs a licence for closed-source repositories. Mantic.sh and Codemod Grep demonstrate search rather than this rewrite; AI coding assistants are not the deterministic structural tools being compared. OpenRewrite JavaScript, checked November 7, 2026; remaining exclusions use October 19 research.

ToolPublic sourcesWhat remains uncertain
ESLintCustom-rule API, CLI dry-run/fatal-parse flags, MIT parser source, releases/security, checked November 7.Custom rule correctness and later apply recovery; no whole-batch rollback.
SemgrepSource, pricing, CLI scan options, releases/security and Homebrew listing, checked November 7.Experimental autofix warning; no safe-write guarantee.
ast-grepRewrite guide, FAQ, TSX catalog, releases, issues/security, October 19 and November 7.No binding/parse gate; safe writes across environments; configuration-loading risk.
codeeditREADME, displayed release and human fixes, checked November 7.Earliest public release, advisory coverage and practical behaviour of maker write guarantees.
Newcomer assessmentGitHub topics/candidate repositories, current Codemod docs and October discovery, checked November 7.Launch provenance/platform gaps block an award; unnamed engine cannot be identified.

None of this establishes speed, accuracy across every TSX shape, behaviour in large repositories or security in practice. Next substantive review is due December 18, 2026, or sooner if maintenance, packaging or security changes. Found something out of date? Contact the editors.

Sources and evidence

Sources were checked October 19, 2026 unless another date is given. Project documentation establishes advertised behaviour, not performance. Issue reports are individual experiences, not failure rates. No vendor supplied access, sponsored this work or influenced the conclusions, and no affiliate links are used.

Additional sources checked November 7, 2026: ESLint custom rules, ESLint releases, typescript-eslint, Codemod pricing, Codemod releases, zmod repository, zmod releases, Semgrep source, Babel documentation, Putout source.

Additional sources checked November 7, 2026: ESLint CLI, typescript-eslint source, Semgrep CLI, codeedit, Threadmoth, nodeseek.

No product screenshot is shown. The downloadable files are an original synthetic fixture from an earlier hands-on check on October 1, 2026, made before we adopted our research-only method. Download the earlier check fixture, transform and expected patch.