{"id":"389b5353-87c0-4367-9384-71035657f18d","shortId":"WQEH5W","kind":"skill","title":"filesystem-context","tagline":"Use for file-based context management, dynamic context discovery, and reducing context window bloat. Offload context to files for just-in-time loading.","description":"# Filesystem-Based Context Engineering\n\nThe filesystem provides a single interface through which agents can flexibly store, retrieve, and update an effectively unlimited amount of context. This pattern addresses the fundamental constraint that context windows are limited while tasks often require more information than fits in a single window.\n\nThe core insight is that files enable dynamic context discovery: agents pull relevant context on demand rather than carrying everything in the context window. This contrasts with static context, which is always included regardless of relevance.\n\n## When to Use\nActivate this skill when:\n- Tool outputs are bloating the context window\n- Agents need to persist state across long trajectories\n- Sub-agents must share information without direct message passing\n- Tasks require more context than fits in the window\n- Building agents that learn and update their own instructions\n- Implementing scratch pads for intermediate results\n- Terminal outputs or logs need to be accessible to agents\n\n## Core Concepts\n\nContext engineering can fail in four predictable ways. First, when the context an agent needs is not in the total available context. Second, when retrieved context fails to encapsulate needed context. Third, when retrieved context far exceeds needed context, wasting tokens and degrading performance. Fourth, when agents cannot discover niche information buried in many files.\n\nThe filesystem addresses these failures by providing a persistent layer where agents write once and read selectively, offloading bulk content while preserving the ability to retrieve specific information through search tools.\n\n## Detailed Topics\n\n### The Static vs Dynamic Context Trade-off\n\n**Static Context**\nStatic context is always included in the prompt: system instructions, tool definitions, and critical rules. Static context consumes tokens regardless of task relevance. As agents accumulate more capabilities (tools, skills, instructions), static context grows and crowds out space for dynamic information.\n\n**Dynamic Context Discovery**\nDynamic context is loaded on-demand when relevant to the current task. The agent receives minimal static pointers (names, descriptions, file paths) and uses search tools to load full content when needed.\n\nDynamic discovery is more token-efficient because only necessary data enters the context window. It can also improve response quality by reducing potentially confusing or contradictory information.\n\nThe trade-off: dynamic discovery requires the model to correctly identify when to load additional context. This works well with current frontier models but may fail with less capable models that do not recognize when they need more information.\n\n### Pattern 1: Filesystem as Scratch Pad\n\n**The Problem**\nTool calls can return massive outputs. A web search may return 10k tokens of raw content. A database query may return hundreds of rows. If this content enters the message history, it remains for the entire conversation, inflating token costs and potentially degrading attention to more relevant information.\n\n**The Solution**\nWrite large tool outputs to files instead of returning them directly to the context. The agent then uses targeted retrieval (grep, line-specific reads) to extract only the relevant portions.\n\n**Implementation**\n```python\ndef handle_tool_output(output: str, threshold: int = 2000) -> str:\n    if len(output) < threshold:\n        return output\n    \n    # Write to scratch pad\n    file_path = f\"scratch/{tool_name}_{timestamp}.txt\"\n    write_file(file_path, output)\n    \n    # Return reference instead of content\n    key_summary = extract_summary(output, max_tokens=200)\n    return f\"[Output written to {file_path}. Summary: {key_summary}]\"\n```\n\nThe agent can then use `grep` to search for specific patterns or `read_file` with line ranges to retrieve targeted sections.\n\n**Benefits**\n- Reduces token accumulation over long conversations\n- Preserves full output for later reference\n- Enables targeted retrieval instead of carrying everything\n\n### Pattern 2: Plan Persistence\n\n**The Problem**\nLong-horizon tasks require agents to make plans and follow them. But as conversations extend, plans can fall out of attention or be lost to summarization. The agent loses track of what it was supposed to do.\n\n**The Solution**\nWrite plans to the filesystem. The agent can re-read its plan at any point, reminding itself of the current objective and progress. This is sometimes called \"manipulating attention through recitation.\"\n\n**Implementation**\nStore plans in structured format:\n```yaml\n# scratch/current_plan.yaml\nobjective: \"Refactor authentication module\"\nstatus: in_progress\nsteps:\n  - id: 1\n    description: \"Audit current auth endpoints\"\n    status: completed\n  - id: 2\n    description: \"Design new token validation flow\"\n    status: in_progress\n  - id: 3\n    description: \"Implement and test changes\"\n    status: pending\n```\n\nThe agent reads this file at the start of each turn or when it needs to re-orient.\n\n### Pattern 3: Sub-Agent Communication via Filesystem\n\n**The Problem**\nIn multi-agent systems, sub-agents typically report findings to a coordinator agent through message passing. This creates a \"game of telephone\" where information degrades through summarization at each hop.\n\n**The Solution**\nSub-agents write their findings directly to the filesystem. The coordinator reads these files directly, bypassing intermediate message passing. This preserves fidelity and reduces context accumulation in the coordinator.\n\n**Implementation**\n```\nworkspace/\n  agents/\n    research_agent/\n      findings.md        # Research agent writes here\n      sources.jsonl      # Source tracking\n    code_agent/\n      changes.md         # Code agent writes here\n      test_results.txt   # Test output\n  coordinator/\n    synthesis.md         # Coordinator reads agent outputs, writes synthesis\n```\n\nEach agent operates in relative isolation but shares state through the filesystem.\n\n### Pattern 4: Dynamic Skill Loading\n\n**The Problem**\nAgents may have many skills or instruction sets, but most are irrelevant to any given task. Stuffing all instructions into the system prompt wastes tokens and can confuse the model with contradictory or irrelevant guidance.\n\n**The Solution**\nStore skills as files. Include only skill names and brief descriptions in static context. The agent uses search tools to load relevant skill content when the task requires it.\n\n**Implementation**\nStatic context includes:\n```\nAvailable skills (load with read_file when relevant):\n- database-optimization: Query tuning and indexing strategies\n- api-design: REST/GraphQL best practices\n- testing-strategies: Unit, integration, and e2e testing patterns\n```\n\nAgent loads `skills/database-optimization/SKILL.md` only when working on database tasks.\n\n### Pattern 5: Terminal and Log Persistence\n\n**The Problem**\nTerminal output from long-running processes accumulates rapidly. Copying and pasting output into agent input is manual and inefficient.\n\n**The Solution**\nSync terminal output to files automatically. The agent can then grep for relevant sections (error messages, specific commands) without loading entire terminal histories.\n\n**Implementation**\nTerminal sessions are persisted as files:\n```\nterminals/\n  1.txt    # Terminal session 1 output\n  2.txt    # Terminal session 2 output\n```\n\nAgents query with targeted grep:\n```bash\ngrep -A 5 \"error\" terminals/1.txt\n```\n\n### Pattern 6: Learning Through Self-Modification\n\n**The Problem**\nAgents often lack context that users provide implicitly or explicitly during interactions. Traditionally, this requires manual system prompt updates between sessions.\n\n**The Solution**\nAgents write learned information to their own instruction files. Subsequent sessions load these files, incorporating learned context automatically.\n\n**Implementation**\nAfter user provides preference:\n```python\ndef remember_preference(key: str, value: str):\n    preferences_file = \"agent/user_preferences.yaml\"\n    prefs = load_yaml(preferences_file)\n    prefs[key] = value\n    write_yaml(preferences_file, prefs)\n```\n\nSubsequent sessions include a step to load user preferences if the file exists.\n\n**Caution**\nThis pattern is still emerging. Self-modification requires careful guardrails to prevent agents from accumulating incorrect or contradictory instructions over time.\n\n### Filesystem Search Techniques\n\nModels are specifically trained to understand filesystem traversal. The combination of `ls`, `glob`, `grep`, and `read_file` with line ranges provides powerful context discovery:\n\n- `ls` / `list_dir`: Discover directory structure\n- `glob`: Find files matching patterns (e.g., `**/*.py`)\n- `grep`: Search file contents for patterns, returns matching lines\n- `read_file` with ranges: Read specific line ranges without loading entire files\n\nThis combination often outperforms semantic search for technical content (code, API docs) where semantic meaning is sparse but structural patterns are clear.\n\nSemantic search and filesystem search work well together: semantic search for conceptual queries, filesystem search for structural and exact-match queries.\n\n## Practical Guidance\n\n### When to Use Filesystem Context\n\n**Use filesystem patterns when:**\n- Tool outputs exceed 2000 tokens\n- Tasks span multiple conversation turns\n- Multiple agents need to share state\n- Skills or instructions exceed what fits comfortably in system prompt\n- Logs or terminal output need selective querying\n\n**Avoid filesystem patterns when:**\n- Tasks complete in single turns\n- Context fits comfortably in window\n- Latency is critical (file I/O adds overhead)\n- Simple model incapable of filesystem tool use\n\n### File Organization\n\nStructure files for discoverability:\n```\nproject/\n  scratch/           # Temporary working files\n    tool_outputs/    # Large tool results\n    plans/           # Active plans and checklists\n  memory/            # Persistent learned information\n    preferences.yaml # User preferences\n    patterns.md      # Learned patterns\n  skills/            # Loadable skill definitions\n  agents/            # Sub-agent workspaces\n```\n\nUse consistent naming conventions. Include timestamps or IDs in scratch files for disambiguation.\n\n### Token Accounting\n\nTrack where tokens originate:\n- Measure static vs dynamic context ratio\n- Monitor tool output sizes before and after offloading\n- Track how often dynamic context is actually loaded\n\nOptimize based on measurements, not assumptions.\n\n## Examples\n\n**Example 1: Tool Output Offloading**\n```\nInput: Web search returns 8000 tokens\nBefore: 8000 tokens added to message history\nAfter: \n  - Write to scratch/search_results_001.txt\n  - Return: \"[Results in scratch/search_results_001.txt. Key finding: API rate limit is 1000 req/min]\"\n  - Agent greps file when needing specific details\nResult: ~100 tokens in context, 8000 tokens accessible on demand\n```\n\n**Example 2: Dynamic Skill Loading**\n```\nInput: User asks about database indexing\nStatic context: \"database-optimization: Query tuning and indexing\"\nAgent action: read_file(\"skills/database-optimization/SKILL.md\")\nResult: Full skill loaded only when relevant\n```\n\n**Example 3: Chat History as File Reference**\n```\nTrigger: Context window limit reached, summarization required\nAction: \n  1. Write full history to history/session_001.txt\n  2. Generate summary for new context window\n  3. Include reference: \"Full history in history/session_001.txt\"\nResult: Agent can search history file to recover details lost in summarization\n```\n\n## Guidelines\n\n1. Write large outputs to files; return summaries and references to context\n2. Store plans and state in structured files for re-reading\n3. Use sub-agent file workspaces instead of message chains\n4. Load skills dynamically rather than stuffing all into system prompt\n5. Persist terminal and log output as searchable files\n6. Combine grep/glob with semantic search for comprehensive discovery\n7. Organize files for agent discoverability with clear naming\n8. Measure token savings to validate filesystem patterns are effective\n9. Implement cleanup for scratch files to prevent unbounded growth\n10. Guard self-modification patterns with validation\n\n## Integration\n\nThis skill connects to:\n\n- context-optimization - Filesystem offloading is a form of observation masking\n- memory-systems - Filesystem-as-memory is a simple memory layer\n- multi-agent-patterns - Sub-agent file workspaces enable isolation\n- context-compression - File references enable lossless \"compression\"\n- tool-design - Tools should return file references for large outputs\n\n## References\n\nInternal reference:\n- Implementation Patterns - Detailed pattern implementations\n\nRelated skills in this collection:\n- context-optimization - Token reduction techniques\n- memory-systems - Persistent storage patterns\n- multi-agent-patterns - Agent coordination\n\nExternal resources:\n- LangChain Deep Agents: How agents can use filesystems for context engineering\n- Cursor: Dynamic context discovery patterns\n- Anthropic: Agent Skills specification\n\n---\n\n## Skill Metadata\n\n**Created**: 2026-01-07\n**Last Updated**: 2026-01-07\n**Author**: Agent Skills for Context Engineering Contributors\n**Version**: 1.0.0\n\n## Limitations\n- Use this skill only when the task clearly matches the scope described above.\n- Do not treat the output as a substitute for environment-specific validation, testing, or expert review.\n- Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.","tags":["filesystem","context","antigravity","awesome","skills","sickn33","agent-skills","agentic-skills","ai-agent-skills","ai-agents","ai-coding","ai-workflows"],"capabilities":["skill","source-sickn33","skill-filesystem-context","topic-agent-skills","topic-agentic-skills","topic-ai-agent-skills","topic-ai-agents","topic-ai-coding","topic-ai-workflows","topic-antigravity","topic-antigravity-skills","topic-claude-code","topic-claude-code-skills","topic-codex-cli","topic-codex-skills"],"categories":["antigravity-awesome-skills"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/sickn33/antigravity-awesome-skills/filesystem-context","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add sickn33/antigravity-awesome-skills","source_repo":"https://github.com/sickn33/antigravity-awesome-skills","install_from":"skills.sh"}},"qualityScore":"0.700","qualityRationale":"deterministic score 0.70 from registry signals: · indexed on github topic:agent-skills · 34793 github stars · SKILL.md body (13,532 chars)","verified":false,"liveness":"unknown","lastLivenessCheck":null,"agentReviews":{"count":0,"score_avg":null,"cost_usd_avg":null,"success_rate":null,"latency_p50_ms":null,"narrative_summary":null,"summary_updated_at":null},"enrichmentModel":"deterministic:skill-github:v1","enrichmentVersion":1,"enrichedAt":"2026-04-24T00:50:56.507Z","embedding":null,"createdAt":"2026-04-18T21:37:12.507Z","updatedAt":"2026-04-24T00:50:56.507Z","lastSeenAt":"2026-04-24T00:50:56.507Z","tsv":"'-01':1798,1803 '-07':1799,1804 '1':426,708,1053,1452,1549,1582 '1.0.0':1813 '1.txt':1050 '10':1675 '100':1493 '1000':1483 '10k':444 '2':614,717,1058,1503,1555,1594 '2.txt':1055 '200':561 '2000':524,1305 '2026':1797,1802 '3':728,756,1535,1562,1606 '4':873,1617 '5':990,1068,1628 '6':1072,1637 '7':1646 '8':1655 '8000':1460,1463,1497 '9':1665 'abil':260 'access':177,1499 'account':1417 'accumul':305,596,825,1004,1179 'across':133 'action':1523,1548 'activ':117,1380 'actual':1442 'ad':1465 'add':1354 'addit':400 'address':57,239 'agent':42,88,128,138,156,179,195,228,248,304,338,498,573,624,647,665,737,759,768,772,779,801,831,833,836,843,846,856,861,879,931,980,1011,1026,1060,1080,1103,1177,1313,1398,1401,1485,1522,1570,1610,1650,1713,1717,1768,1770,1776,1778,1791,1806 'agent/user_preferences.yaml':1136 'also':374 'alway':109,283 'amount':52 'anthrop':1790 'api':966,1257,1479 'api-design':965 'ask':1509,1847 'assumpt':1449 'attent':476,640,688 'audit':710 'auth':712 'authent':701 'author':1805 'automat':1024,1120 'avail':202,949 'avoid':1335 'base':8,31,1445 'bash':1065 'benefit':593 'best':969 'bloat':18,124 'boundari':1855 'brief':925 'build':155 'bulk':255 'buri':233 'bypass':815 'call':434,686 'cannot':229 'capabl':307,414 'care':1173 'carri':96,611 'caution':1163 'chain':1616 'chang':733 'changes.md':844 'chat':1536 'checklist':1383 'clarif':1849 'cleanup':1667 'clear':1268,1653,1822 'code':842,845,1256 'collect':1753 'combin':1198,1248,1638 'comfort':1324,1346 'command':1036 'communic':760 'complet':715,1340 'comprehens':1644 'compress':1724,1729 'concept':181 'conceptu':1280 'confus':381,906 'connect':1686 'consist':1404 'constraint':60 'consum':297 'content':256,354,448,459,553,939,1229,1255 'context':3,9,12,16,20,32,54,62,86,91,100,106,126,149,182,193,203,207,212,216,220,274,279,281,296,312,322,325,370,401,496,824,929,947,1083,1119,1211,1297,1344,1426,1440,1496,1514,1542,1560,1593,1689,1723,1755,1783,1787,1809 'context-compress':1722 'context-optim':1688,1754 'contradictori':383,910,1182 'contrast':103 'contributor':1811 'convent':1406 'convers':469,599,633,1310 'coordin':778,810,828,852,854,1771 'copi':1006 'core':79,180 'correct':395 'cost':472 'creat':784,1796 'criteria':1858 'critic':293,1351 'crowd':315 'current':335,406,679,711 'cursor':1785 'data':367 'databas':450,958,987,1511,1516 'database-optim':957,1515 'deep':1775 'def':516,1127 'definit':291,1397 'degrad':224,475,791 'demand':93,330,1501 'describ':1826 'descript':344,709,718,729,926 'design':719,967,1732 'detail':268,1491,1577,1746 'dir':1215 'direct':143,493,805,814 'directori':1217 'disambigu':1415 'discov':230,1216 'discover':1368,1651 'discoveri':13,87,323,358,390,1212,1645,1788 'doc':1258 'dynam':11,85,273,319,321,324,357,389,874,1425,1439,1504,1620,1786 'e.g':1224 'e2e':977 'effect':50,1664 'effici':363 'emerg':1168 'enabl':84,606,1720,1727 'encapsul':210 'endpoint':713 'engin':33,183,1784,1810 'enter':368,460 'entir':468,1039,1245 'environ':1838 'environment-specif':1837 'error':1033,1069 'everyth':97,612 'exact':1288 'exact-match':1287 'exampl':1450,1451,1502,1534 'exceed':218,1304,1321 'exist':1162 'expert':1843 'explicit':1089 'extend':634 'extern':1772 'extract':509,556 'f':538,563 'fail':185,208,411 'failur':241 'fall':637 'far':217 'fidel':821 'file':7,22,83,236,345,488,536,545,546,567,585,740,813,919,954,1023,1048,1111,1116,1135,1141,1148,1161,1205,1221,1228,1236,1246,1352,1363,1366,1373,1413,1487,1525,1539,1574,1587,1601,1611,1636,1648,1670,1718,1725,1736 'file-bas':6 'filesystem':2,30,35,238,427,663,762,808,871,1186,1195,1272,1282,1296,1299,1336,1360,1661,1691,1703,1781 'filesystem-as-memori':1702 'filesystem-bas':29 'filesystem-context':1 'find':775,804,1220,1478 'findings.md':834 'first':190 'fit':73,151,1323,1345 'flexibl':44 'flow':723 'follow':629 'form':1695 'format':696 'four':187 'fourth':226 'frontier':407 'full':353,601,1528,1551,1565 'fundament':59 'game':786 'generat':1556 'given':893 'glob':1201,1219 'grep':503,577,1029,1064,1066,1202,1226,1486 'grep/glob':1639 'grow':313 'growth':1674 'guard':1676 'guardrail':1174 'guidanc':913,1292 'guidelin':1581 'handl':517 'histori':463,1041,1468,1537,1552,1566,1573 'history/session_001.txt':1554,1568 'hop':796 'horizon':621 'hundr':454 'i/o':1353 'id':707,716,727,1410 'identifi':396 'implement':164,514,691,730,829,945,1042,1121,1666,1744,1748 'implicit':1087 'improv':375 'incap':1358 'includ':110,284,920,948,1152,1407,1563 'incorpor':1117 'incorrect':1180 'index':963,1512,1521 'ineffici':1016 'inflat':470 'inform':71,141,232,264,320,384,424,480,790,1106,1387 'input':1012,1456,1507,1852 'insight':80 'instead':489,551,609,1613 'instruct':163,289,310,885,897,1110,1183,1320 'int':523 'integr':975,1683 'interact':1091 'interfac':39 'intermedi':168,816 'intern':1742 'irrelev':890,912 'isol':865,1721 'just-in-tim':24 'key':554,570,1130,1143,1477 'lack':1082 'langchain':1774 'larg':484,1376,1584,1739 'last':1800 'latenc':1349 'later':604 'layer':246,1710 'learn':158,1073,1105,1118,1386,1392 'len':527 'less':413 'limit':65,1481,1544,1814 'line':505,587,1207,1234,1241 'line-specif':504 'list':1214 'load':28,327,352,399,876,936,951,981,1038,1114,1138,1156,1244,1443,1506,1530,1618 'loadabl':1395 'log':173,993,1328,1632 'long':134,598,620,1001 'long-horizon':619 'long-run':1000 'lose':648 'lossless':1728 'lost':643,1578 'ls':1200,1213 'make':626 'manag':10 'mani':235,882 'manipul':687 'manual':1014,1095 'mask':1698 'massiv':437 'match':1222,1233,1289,1823 'max':559 'may':410,442,452,880 'mean':1261 'measur':1422,1447,1656 'memori':1384,1700,1705,1709,1761 'memory-system':1699,1760 'messag':144,462,781,817,1034,1467,1615 'metadata':1795 'minim':340 'miss':1860 'model':393,408,415,908,1189,1357 'modif':1077,1171,1679 'modul':702 'monitor':1428 'multi':767,1712,1767 'multi-ag':766 'multi-agent-pattern':1711,1766 'multipl':1309,1312 'must':139 'name':343,541,923,1405,1654 'necessari':366 'need':129,174,196,211,219,356,422,750,1314,1332,1489 'new':720,1559 'nich':231 'object':680,699 'observ':1697 'offload':19,254,1435,1455,1692 'often':68,1081,1249,1438 'on-demand':328 'oper':862 'optim':959,1444,1517,1690,1756 'organ':1364,1647 'orient':754 'origin':1421 'outperform':1250 'output':122,171,438,486,519,520,528,531,548,558,564,602,851,857,998,1009,1021,1054,1059,1303,1331,1375,1430,1454,1585,1633,1740,1832 'overhead':1355 'pad':166,430,535 'pass':145,782,818 'past':1008 'path':346,537,547,568 'pattern':56,425,582,613,755,872,979,989,1071,1165,1223,1231,1266,1300,1337,1393,1662,1680,1714,1745,1747,1765,1769,1789 'patterns.md':1391 'pend':735 'perform':225 'permiss':1853 'persist':131,245,616,994,1046,1385,1629,1763 'plan':615,627,635,660,671,693,1379,1381,1596 'point':674 'pointer':342 'portion':513 'potenti':380,474 'power':1210 'practic':970,1291 'predict':188 'pref':1137,1142,1149 'prefer':1125,1129,1134,1140,1147,1158,1390 'preferences.yaml':1388 'preserv':258,600,820 'prevent':1176,1672 'problem':432,618,764,878,996,1079 'process':1003 'progress':682,705,726 'project':1369 'prompt':287,901,1097,1327,1627 'provid':36,243,1086,1124,1209 'pull':89 'py':1225 'python':515,1126 'qualiti':377 'queri':451,960,1061,1281,1290,1334,1518 'rang':588,1208,1238,1242 'rapid':1005 'rate':1480 'rather':94,1621 'ratio':1427 'raw':447 're':668,753,1604 're-ori':752 're-read':667,1603 'reach':1545 'read':252,507,584,669,738,811,855,953,1204,1235,1239,1524,1605 'receiv':339 'recit':690 'recogn':419 'recov':1576 'reduc':15,379,594,823 'reduct':1758 'refactor':700 'refer':550,605,1540,1564,1591,1726,1737,1741,1743 'regardless':111,299 'relat':864,1749 'relev':90,113,302,332,479,512,937,956,1031,1533 'remain':465 'rememb':1128 'remind':675 'report':774 'req/min':1484 'requir':69,147,391,623,943,1094,1172,1547,1851 'research':832,835 'resourc':1773 'respons':376 'rest/graphql':968 'result':169,1378,1474,1492,1527,1569 'retriev':46,206,215,262,502,590,608 'return':436,443,453,491,530,549,562,1232,1459,1473,1588,1735 'review':1844 'row':456 'rule':294 'run':1002 'safeti':1854 'save':1658 'scope':1825 'scratch':165,429,534,539,1370,1412,1669 'scratch/current_plan.yaml':698 'scratch/search_results_001.txt':1472,1476 'search':266,349,441,579,933,1187,1227,1252,1270,1273,1278,1283,1458,1572,1642 'searchabl':1635 'second':204 'section':592,1032 'select':253,1333 'self':1076,1170,1678 'self-modif':1075,1169,1677 'semant':1251,1260,1269,1277,1641 'session':1044,1052,1057,1100,1113,1151 'set':886 'share':140,867,1316 'simpl':1356,1708 'singl':38,76,1342 'size':1431 'skill':119,309,875,883,917,922,938,950,1318,1394,1396,1505,1529,1619,1685,1750,1792,1794,1807,1817 'skill-filesystem-context' 'skills/database-optimization/skill.md':982,1526 'solut':482,658,798,915,1018,1102 'sometim':685 'sourc':840 'source-sickn33' 'sources.jsonl':839 'space':317 'span':1308 'spars':1263 'specif':263,506,581,1035,1191,1240,1490,1793,1839 'start':743 'state':132,868,1317,1598 'static':105,271,278,280,295,311,341,928,946,1423,1513 'status':703,714,724,734 'step':706,1154 'still':1167 'stop':1845 'storag':1764 'store':45,692,916,1595 'str':521,525,1131,1133 'strategi':964,973 'structur':695,1218,1265,1285,1365,1600 'stuf':895,1623 'sub':137,758,771,800,1400,1609,1716 'sub-ag':136,757,770,799,1399,1608,1715 'subsequ':1112,1150 'substitut':1835 'success':1857 'summar':645,793,1546,1580 'summari':555,557,569,571,1557,1589 'suppos':654 'sync':1019 'synthesi':859 'synthesis.md':853 'system':288,769,900,1096,1326,1626,1701,1762 'target':501,591,607,1063 'task':67,146,301,336,622,894,942,988,1307,1339,1821 'technic':1254 'techniqu':1188,1759 'telephon':788 'temporari':1371 'termin':170,991,997,1020,1040,1043,1049,1051,1056,1330,1630 'terminals/1.txt':1070 'test':732,850,972,978,1841 'test_results.txt':849 'testing-strategi':971 'third':213 'threshold':522,529 'time':27,1185 'timestamp':542,1408 'togeth':1276 'token':222,298,362,445,471,560,595,721,903,1306,1416,1420,1461,1464,1494,1498,1657,1757 'token-effici':361 'tool':121,267,290,308,350,433,485,518,540,934,1302,1361,1374,1377,1429,1453,1731,1733 'tool-design':1730 'topic':269 'topic-agent-skills' 'topic-agentic-skills' 'topic-ai-agent-skills' 'topic-ai-agents' 'topic-ai-coding' 'topic-ai-workflows' 'topic-antigravity' 'topic-antigravity-skills' 'topic-claude-code' 'topic-claude-code-skills' 'topic-codex-cli' 'topic-codex-skills' 'total':201 'track':649,841,1418,1436 'trade':276,387 'trade-off':275,386 'tradit':1092 'train':1192 'trajectori':135 'travers':1196 'treat':1830 'trigger':1541 'tune':961,1519 'turn':746,1311,1343 'txt':543 'typic':773 'unbound':1673 'understand':1194 'unit':974 'unlimit':51 'updat':48,160,1098,1801 'use':4,116,348,500,576,932,1295,1298,1362,1403,1607,1780,1815 'user':1085,1123,1157,1389,1508 'valid':722,1660,1682,1840 'valu':1132,1144 'version':1812 'via':761 'vs':272,1424 'wast':221,902 'way':189 'web':440,1457 'well':404,1275 'window':17,63,77,101,127,154,371,1348,1543,1561 'without':142,1037,1243 'work':403,985,1274,1372 'workspac':830,1402,1612,1719 'write':249,483,532,544,659,802,837,847,858,1104,1145,1470,1550,1583 'written':565 'yaml':697,1139,1146","prices":[{"id":"eb92fd20-4a6b-4422-bebf-9d9d599bda77","listingId":"389b5353-87c0-4367-9384-71035657f18d","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"sickn33","category":"antigravity-awesome-skills","install_from":"skills.sh"},"createdAt":"2026-04-18T21:37:12.507Z"}],"sources":[{"listingId":"389b5353-87c0-4367-9384-71035657f18d","source":"github","sourceId":"sickn33/antigravity-awesome-skills/filesystem-context","sourceUrl":"https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/filesystem-context","isPrimary":false,"firstSeenAt":"2026-04-18T21:37:12.507Z","lastSeenAt":"2026-04-24T00:50:56.507Z"}],"details":{"listingId":"389b5353-87c0-4367-9384-71035657f18d","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"sickn33","slug":"filesystem-context","github":{"repo":"sickn33/antigravity-awesome-skills","stars":34793,"topics":["agent-skills","agentic-skills","ai-agent-skills","ai-agents","ai-coding","ai-workflows","antigravity","antigravity-skills","claude-code","claude-code-skills","codex-cli","codex-skills","cursor","cursor-skills","developer-tools","gemini-cli","gemini-skills","kiro","mcp","skill-library"],"license":"mit","html_url":"https://github.com/sickn33/antigravity-awesome-skills","pushed_at":"2026-04-24T00:28:59Z","description":"Installable GitHub library of 1,400+ agentic skills for Claude Code, Cursor, Codex CLI, Gemini CLI, Antigravity, and more. Includes installer CLI, bundles, workflows, and official/community skill collections.","skill_md_sha":"81c3dd33dbdbb88008a154749f6b79db5996afd6","skill_md_path":"skills/filesystem-context/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/filesystem-context"},"layout":"multi","source":"github","category":"antigravity-awesome-skills","frontmatter":{"name":"filesystem-context","description":"Use for file-based context management, dynamic context discovery, and reducing context window bloat. Offload context to files for just-in-time loading."},"skills_sh_url":"https://skills.sh/sickn33/antigravity-awesome-skills/filesystem-context"},"updatedAt":"2026-04-24T00:50:56.507Z"}}