Review local change declarations, release bytes and source records
Inspect Terraform masks, Helm bytes, bookmark provenance, FASTQ quality, CMake artifact paths and PDB SEQRES with complete downloads and explicit limits.
Choose the receiving task
A change declaration, stored release, bookmark export, quality-filter decision, artifact location and declared protein chain answer different questions. Preserve the format semantics and the source needed to verify each handoff.
Terraform: read actions, presence and masks together
The starter contains three resource changes, one drift record and two output changes. Two replacements contribute both a create leg and a delete leg. A deposed object sharing an address remains a separate record. applyable, complete and errored are snapshot flags; they do not establish that apply will succeed.
Sensitive true masks redact the whole marked subtree. Unknown values retain explicit markers, while null and absent stay distinct. RFC6901 paths identify each marked subtree. The number token 9007199254740993 remains exact. Missing masks mean no declaration: unmarked values, addresses and import IDs can still be private.
Raw variables, planned_values, prior_state, configuration and unknown top-field values are omitted from every output. Unknown resource/change fields are named as unreviewed. Seven traditional ordered action forms are supported; other actions and nonempty deferred_changes reject. No Terraform, HCL or provider runs.
Helm: separate encoded, expanded and chart byte counts
The starter decodes a Secret into an exact 565-byte release, three chart byte entries and one hook. release.json retains the number token 9007199254740993. Root field spans measure original UTF-8 JSON value tokens; chart sizes and hashes measure decoded Base64 bytes. These are different measurements and do not predict Kubernetes Secret acceptance.
Select Secret JSON for kind Secret, type helm.sh/release.v1 and data.release; select encoded-release for inner Base64 text. Gzip CRC, truncation, expanded size and UTF-8 are checked. Legacy plain releases are labelled. The complete release, literal manifest and hooks are downloadable; unknown root fields remain unreviewed. Subchart trees are not reconstructed and templates or hooks do not execute.
Bookmarks: inspect chosen records and complete alternatives
Select at least two UTF-8 Netscape exports. The starter merges nine semantic nodes and five bookmarks into seven nodes and four bookmarks: one exact same-folder HREF duplicate and one exact-title folder merge. The same HREF in a different folder stays. First source and record win the chosen title and attributes; every alternate record, tag, icon and description remains in provenance and conflicts.
Attribute order creates no false conflict; decoded attribute names are case-insensitive. Optional empty-folder removal runs after merge and records each removal. Bookmark DD text is preserved, including line breaks. Folder DD, ambiguous structures and duplicate attributes reject. Modern Firefox may ignore DD, so retaining it in this export does not promise full profile recovery. No URL is fetched and no browser profile is written.
FASTQ: apply the inclusive interval to every base
The quality characters ?@ABCDEFGHIJK represent PHRED+33 scores 30–42. Q43 is outside this interval despite exceeding its minimum. The starter keeps one read and drops two. Every record has minimum, maximum, mean, outside count and a reason; scores 0–93 fill exactly 94 histogram bins.
Each nonempty IUPAC DNA/RNA read has four physical lines and equal sequence/quality lengths. A nonempty plus identifier must equal the full header. Quality @ and + are ordinary characters. Consistent LF or CRLF and selected records, including final EOF state, remain byte-exact. Empty whole input is valid; wrapped FASTQ, zero-base reads, Phred64 guessing and paired synchronization are unsupported.
CMake: resolve all artifacts against the top-level build root
Select the index, referenced codemodel and target reply JSON files. The included CMake 3.31.6 configure-only snapshot declares four targets and three artifacts. Utility targets remain visible with no-declared-artifacts status. Every configuration, executable/library/PDB artifact and declared dependency is preserved.
Relative artifact paths resolve lexically against codemodel.paths.build. target.paths.build and nameOnDisk remain visible declarations and do not replace that root. Choose POSIX or Windows rules explicitly. Drive and UNC paths have fixed vectors; drive-relative C:foo is unsupported. Dependencies are declared or dangling references within their configuration.
Other reply objects/files and unknown target fields are uninterpreted. Missing references, mismatched target ID/name, duplicate basenames and unsupported versions reject atomically. No disk existence, build success, CMake command or executable is inferred or run.
PDB: export declared chains without coordinate inference
The starter declares 21 residues on A and MSE/UNK/ALA on B. Unique MODRES maps MSE to MET, giving MXA with UNK explicitly unknown. Select all chains or one printable ASCII character, including space. chain-u0020 reversibly represents blank; all 95 printable ASCII chain IDs stay distinct.
Legacy ASCII PDB 3.3 protein SEQRES uses fixed three-character slots, continuous serials and constant numRes matching the total. Twenty standard amino acids plus ASX/GLX/UNK and unique MODRES parents are supported. Other protein residue names become X with original name, parent and position. Nucleotides, mmCIF and non-ASCII input reject. Missing SEQRES yields no declared sequence, without ATOM fallback or biological validation.
Download the complete handoff and recover deliberately
Input, structural and full-output budgets apply together; maximum counts do not promise every combined maximum fits. Over-budget or malformed input rejects without partial downloads. The table previews 200 rows, with each cell showing at most 2,000 characters. Above 20,000 report characters, text and copy contain a labelled preview; full downloads remain complete.
Keep original sources and selected parameters with the handoff. Download while the tab is open. Source URLs and paths are data; no source command, browser profile, sequence program or external service is executed.
| Task | Input | Structure / complete output |
|---|---|---|
| Terraform | 1 file/paste 5MiB | resource/drift/output each10k; JSON100k; depth64; output20MiB |
| Helm | 1 file/paste5MiB | expanded8MiB; chart10k; aggregateJSON100k; depth64; output20MiB |
| Bookmarks | 2–20 files; each2MiB; total10MiB | semantic nodes50k; bookmarks20k; folders depth64; output20MiB |
| FASTQ | 1 file/paste10MiB | records100k; bases4M; output40MiB |
| CMake | up to2000 files; each1MiB; total10MiB | targets10k; artifacts20k; dependencies50k; aggregateJSON500k; depth64; output20MiB |
| PDB | 1 file/paste10MiB | lines200k; chains95; residues100k; output20MiB |
- Terraform: correct action/mask schemas or export a supported plan; unmarked data is not automatically private.
- Helm: confirm the explicit encoding mode and export complete canonical Base64 / UTF-8 JSON.
- Bookmarks: inspect same-parent folder/HREF identity and all alternate metadata before importing.
- FASTQ: correct lengths/newlines and the intended inclusive quality interval.
- CMake: include referenced basenames, select one index and choose the snapshot path convention.
- PDB: correct fixed slots/serials/counts or choose an existing chain; use a separate nucleotide/mmCIF workflow.
- After cancellation, rerun the selected valid sources and download complete results.
References
- Terraform long-plan review question
First-person task demand; the supported scope is defined in this article.
- Terraform JSON format
Read2026-10-07; format/library correctness reference, not demand evidence.
- Oversized Helm Secret question
First-person task demand; the supported scope is defined in this article.
- Helm storage decoding
Read2026-10-07; format/library correctness reference, not demand evidence.
- Merge bookmark exports question
First-person task demand; the supported scope is defined in this article.
- Firefox bookmark importer
Read2026-10-07; format/library correctness reference, not demand evidence.
- Exact whole-read quality question
First-person task demand; the supported scope is defined in this article.
- Biopython FASTQ reader
Read2026-10-07; format/library correctness reference, not demand evidence.
- CMake complete artifact paths question
First-person task demand; the supported scope is defined in this article.
- CMake File API
Read2026-10-07; format/library correctness reference, not demand evidence.
- Specific declared PDB chain question
First-person task demand; the supported scope is defined in this article.
- wwPDB SEQRES fields
Read2026-10-07; format/library correctness reference, not demand evidence.
Tools in this category
Expand a tool to see its steps, options and supported formats, then open its workspace.
Terraform plan declaration reviewReview ordered Terraform plan actions, deposed copies, drift and output changes with recursive supplied sensitive and unknown masks.
Read a long exported plan before sharing its change list. Keep replacement order and snapshot flags explicit while withholding marked values and unrelated raw plan sections.
Steps
- Paste terraform show -json output or select its exported JSON file.
- Review action order,deposed identity,drift/output sections and supplied mask pointers.
- Download complete masked JSON/CSV; retain the original privately when further review needs omitted sections.
Available options
- Protect CSV formula-like text
- On by default
Capabilities and limits
- One UTF-8 file/paste5MiB;10,000 resource changes,10,000 drift changes,10,000 output changes,100,000 JSON values,depth64. Complete downloads20MiB; budgets apply together and failures return no partial report.
- format_version1.x with resource_changes is required; state-only JSON,HCL and binary plans unsupported. Support only no-op/read/create/update/delete/delete→create/create→delete. Nonempty deferred_changes and other actions reject. Address+deposed identifies copies; the same address with different deposed keys remains.
- before/after preserve absent/null/present separately. Sensitive true subtrees become explicit redacted-sensitive markers; after_unknown true becomes unknown, including declared fields absent from partial after. Mask arrays must match shape. RFC6901 true-mask pointers are complete; missing optional masks mean no declaration, not automatic detection.
- Retained declared numeric tokens,replace_paths,previous_address,index,action_reason and importing declarations remain exact. Raw variables/planned_values/prior_state/configuration and other unreviewed top-field values are not exported; unknown resource/change field names are listed without their values.
- Unmarked values, addresses and import IDs can remain private. Flags/actions describe only the snapshot, not apply success,safety or real impact. Preview200 rows/cell2000; above20,000 report characters, copy is a labelled preview; full JSON/CSV remain complete. No Terraform/provider/HCL command or request runs.
Helm release Secret inspectorDecode a local Helm3 release Secret, preserve the exact release and manifest, and compare raw JSON field spans with decoded chart file sizes and hashes.
Investigate an oversized stored release from its exported Secret. See what was actually serialized, without rendering a chart or connecting to Kubernetes.
Steps
- Paste an exported Helm3 Secret JSON or select the explicit encoded-release mode.
- Compare the raw-field byte spans and decoded chart file sizes; review unreviewed fields and the encoding layer.
- Download the exact release, literal manifest, hook declarations and complete inventory.
Available options
- Input mode
- Secret JSON · Encoded release
- Protect CSV formula-like text
- On by default
Capabilities and limits
- One UTF-8 file or paste up to 5 MiB. Release expansion up to 8 MiB; 10,000 chart byte entries, 100,000 JSON values across envelope and release, 64 levels. Complete downloads up to 20 MiB; over-budget or invalid data rejects atomically.
- Select Secret JSON for Kubernetes kind Secret/type helm.sh/release.v1/data.release, or encoded release for the inner Helm Base64 text. Base64 is canonical and unwrapped. Gzip is streamed with CRC/truncation checks; legacy uncompressed releases are identified. YAML input is unsupported.
- release.json preserves the complete decoded UTF-8 bytes, including large number tokens. manifest.yaml preserves its literal string. hooks.json preserves hook declarations. Templates/files/schema are measured and SHA256-hashed as original decoded bytes; templates are never executed.
- Field spans are half-open byte offsets into release.json and count each original JSON value token. They are separate from decoded chart bytes and gzip payload size; sums do not predict compressed contribution or Kubernetes acceptance. Only the serialized root chart is inspected. No subchart dependency tree is reconstructed.
- Release/config/manifest can contain private values. Exports remain local. Unknown root fields are retained and labelled unreviewed. Preview is first 200 rows; cells over 2,000 characters are clipped. Above 20,000 report characters, copying gives the labelled preview; complete downloads remain available.
Browser bookmark export mergerMerge UTF-8 Netscape bookmark exports by exact folder title and same-folder HREF, preserving alternate metadata and complete source provenance.
Combine several exported bookmark files without installing a profile editor. Keep the chosen hierarchy, every duplicate record and conflicting tags/descriptions checkable.
Steps
- Choose at least two UTF-8 browser bookmark HTML exports.
- Review chosen hierarchy,duplicate metadata and the exact same-folder URL policy; optionally remove empty folders.
- Download bookmarks.html and the complete provenance/conflict report before importing through your browser.
Available options
- Remove empty folders after merge
- Off by default
- Protect CSV formula-like text
- On by default
Capabilities and limits
- Select2–20 UTF-8 bookmark HTML exports, each2MiB,total10MiB; no paste. Across all inputs50,000 folder/bookmark/separator semantic nodes,20,000 bookmarks,folder depth64; synthetic file roots excluded. Complete output20MiB,atomic over-budget failure.
- Only exact-title folders within the same parent merge. Only exact parser-decoded HREF strings within the same output folder deduplicate. The same URL in different folders stays; first-file/first-record precedence is stable. No URL normalization or fetching.
- Chosen titles,legal attributes and bookmark DD descriptions remain in importable Netscape HTML. Every input semantic record and alternate metadata is preserved in provenance; duplicate bookmarks also retain chosen/alternate snapshots in conflicts. Remove-empty-folders is optional and runs after merge, with removal provenance.
- Folder DD,orphan/duplicate DD,duplicate ambiguous HTML attributes,unrecognized semantic elements or ambiguous/unclosed DL/H3/A structure reject. META encoding must match UTF-8. Attribute names are HTML case-insensitive; decoded text/values are preserved, not original HTML bytes/comments/header formatting. Modern Firefox DD import is not guaranteed.
- Results contain literal URL strings with no clickable result links or DOM injection. HTML is a downloaded import file, never executed by this tool. No profile write/sync/network. Table200 rows/cell2000; text/copy above20,000 characters is a labelled preview; full JSON/CSV/HTML remain complete.
FASTQ read quality filterFilter strict four-line single-end PHRED+33 FASTQ by an inclusive all-base score interval, retaining selected original record bytes and complete decisions.
Keep or drop whole reads using an explicit quality interval. The default 30–42 matches the quality characters ? through K; legal @ and + characters stay quality data.
Steps
- Open a four-line PHRED+33 FASTQ and choose the inclusive score interval.
- Inspect kept/dropped counts, score histograms and complete per-record reasons.
- Download the byte-preserving filtered FASTQ and all decision rows.
Available options
- Inclusive minimum
- 30
- Inclusive maximum
- 42
- Protect CSV formula-like text
- On by default
Capabilities and limits
- One UTF-8 file or paste up to 10 MiB, 100,000 records and 4,000,000 total bases. Complete downloads up to 40 MiB. Empty whole files produce zero records; nonempty records require at least one base.
- Strict four lines per record: nonempty @identifier, IUPAC DNA/RNA sequence, + with an optional identical full identifier, and equal-length quality. PHRED+33 ASCII ! through ~ scores 0–93. Wrapped FASTQ, Phred64 guessing and paired synchronization are unsupported.
- Every base must fall inside the inclusive minimum/maximum interval, default 30–42. A read containing Q43 is dropped too. Reports include minimum/maximum/mean, outside count, kept/dropped reason and two complete 94-bin score histograms. Means are arithmetic score averages.
- Use only LF or only CRLF within one file. Mixed conventions and lone CR reject. A final quality line without a newline is legal. Each kept entire source record preserves header, plus line, bytes, ordering and EOF state. Byte spans are half-open UTF-8 offsets.
- Table preview shows 200 decisions with long cells clipped to 2,000 characters. Above 20,000 report characters copying gives an explicit preview. Complete FASTQ/JSON/CSV remain downloadable; invalid or over-budget input returns no partial filtered file. No biological or medical interpretation.
CMake File API artifact inventoryJoin local codemodel2 reply files across configurations, resolving every declared artifact against the top-level build root with full dependencies.
Find full declared output paths without guessing from nameOnDisk or a target’s subdirectory. Read the exported reply snapshot locally, without compiling or running project commands.
Steps
- Select index,codemodel and referenced target reply JSON files.
- Choose the index when ambiguous and the snapshot’s POSIX/Windows path mode.
- Review every artifact,utility status and dependency; download complete JSON/CSV for a separate packaging workflow.
Available options
- Selected index basename (required when multiple)
- Enter as needed
- Path mode
- POSIX · Windows
- Protect CSV formula-like text
- On by default
Capabilities and limits
- Select up to2,000 distinct JSON basenames,each1MiB,total10MiB; no paste. Targets across configurations10,000,artifacts20,000,dependencies50,000,aggregate JSON values500,000,depth64. Complete downloads20MiB; combined budgets apply atomically.
- Select an index basename when several indexes exist; one is chosen automatically only when unique. References must be single legal basenames in the collection. Missing/traversal/ambiguous references,id/name mismatch and duplicate target IDs within a configuration reject; IDs reused across configurations remain.
- Only codemodel major2/minor0..7 snapshots are supported. Every configuration,target and artifact remains; no artifact is an explicit no-declared-artifacts status. nameOnDisk does not replace missing artifacts. Unknown reply objects/files/target fields are listed as uninterpreted.
- Relative artifact paths resolve against codemodel.paths.build,never target.paths.build. Choose POSIX or Windows lexical mode; preserve absolute paths,drive and UNC anchors while normalizing dots. Windows drive-relative C:foo is ambiguous and unsupported; dependencies outside a config remain dangling declarations.
- No path stat/build/command/CMakeLists execution; command fragments remain uninterpreted. Declared paths do not prove a file exists or a build succeeded. Preview200 rows/cell2000; >20,000 report characters gives labelled text/copy preview; full JSON/CSV remain complete.
PDB declared chain sequence exporterExport legacy PDB protein SEQRES declarations as chain-labelled FASTA, with fixed-column residue provenance, explicit MODRES mappings and unknown positions.
Extract the sequence declared for a selected chain. SEQRES declarations can differ from residues with coordinates; the report keeps those meanings separate.
Steps
- Open an ASCII PDB with SEQRES records and select all chains or one exact chain character.
- Compare input/selected counts and inspect unknown or MODRES-mapped residue positions.
- Download declared chain FASTA and complete source-position reports.
Available options
- Chain selection
- All chains · One exact character
- Exact chain; space for blank
- A
- Protect CSV formula-like text
- On by default
Capabilities and limits
- One ASCII legacy text file or paste up to10 MiB; 200,000 lines, 95 printable ASCII chain IDs including blank, and100,000 residues. Complete downloads up to20 MiB. Each chain numRes uses its four-digit field and must match all collected residues.
- Fixed PDB columns only: serial starts1 and increments for each chain, count stays constant, up to13 populated three-character slots per SEQRES line. Invalid separators, gaps, short codes, extra residue fields or count mismatches reject. Nucleotide profiles, mmCIF and non-ASCII input are unsupported.
- Maps20 standard amino acids plus ASX/GLX/UNK. Unique chain+residue-name MODRES parent mappings are used; conflicting parents reject. Other uppercase/digit three-character protein names becomeX with original name, parent, one-based position, source line and column. Unknowns remain visible.
- Select all chains or one exact character, including a space for blank. A missing selected chain rejects. FASTA chain-uXXXX identifiers reversibly encode the original ASCII code; blank is chain-u0020. Selected reports still show complete input chain/residue counts. No SEQRES means no-declared-sequence; ATOM is never substituted.
- FASTA, JSON and CSV include every selected residue. Table preview first 200 rows, cells up to2,000 characters; copying reports above20,000 characters gives a labelled preview. This converts declarations and does not validate observed coordinates, infer missing residues or establish biological facts.